LED light source control method for stabilizing fluorescence emission intensity

Abstract

The invention discloses an LED light source control method for stabilizing fluorescence emission intensity, and the method employs a fluorescence emission intensity error parameter e (t) obtained through calculation based on an emission spectrum prediction model of an LED light source as the feedback of PID control, and the emission spectrum prediction model fully considers the emission spectrum change of the LED light source caused by a junction temperature T and a driving current I; a driving current error compensation parameter is generated in combination with an incremental PID control algorithm, so that the driving current I of the LED light source is regulated and controlled to compensate the fluorescence emission intensity change caused by excitation light source factors such as output power attenuation and spectrum drift of the LED light source; the accuracy and the stability of the fluorescence emission intensity of fluorescence emitted by the fluorescent dye excited by the LED light source are ensured, and the error fluctuation between the actual fluorescence emission intensity and the target fluorescence emission intensity Ftarget can be accurately and stably maintained within 0.1% through tests.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of fluorescence illumination imaging, and particularly relates to an LED light source control method for stabilizing fluorescence emission intensity. BACKGROUND

[0002] Fluorescence microscopy is a core technology in life science and biomedical research, and the accuracy of its quantitative results highly depends on the spectral and optical power stability of the excitation light source during long-time operation. With the development of solid-state lighting technology, light-emitting diodes (LEDs) have gradually replaced traditional mercury lamps and xenon lamps and become the mainstream fluorescence excitation light source due to their long service life, fast response speed, strong wavelength spectrum selectivity, and low power consumption.

[0003] However, as a semiconductor light-emitting device, the spectral characteristics of an LED are significantly affected by the dynamic changes of the junction temperature and the driving current. Typical performances include red shift of the peak wavelength with temperature rise, broadening of the half-peak width with current change, and overall light intensity decay. These changes in spectral morphology will directly lead to fluctuations in the effective excitation efficiency of the fluorescence molecules, thereby introducing instability in the fluorescence emission intensity and affecting the long-time imaging of living cells, high-throughput screening, and quantitative FRET (Fluorescence Resonance Energy Transfer) measurement, and other application scenarios with extremely high stability requirements.

[0004] In addition, the influence of temperature-induced LED spectral drift on fluorescence excitation efficiency has obvious direction dependence. When the initial emission peak wavelength of the LED is located on the long-wavelength side of the fluorescence probe excitation spectrum, the spectral red shift makes the emission spectrum further deviate from the excitation peak position, resulting in a decrease in effective excitation power, and the decay rate of the fluorescence signal may exceed the decline amplitude of the LED output optical power; on the contrary, when the initial peak wavelength of the LED is located on the short-wavelength side of the excitation spectrum, the red shift may make the emission spectrum close to the excitation peak, so that the effective excitation power increases, and a temporary enhancement of the fluorescence signal occurs. Therefore, the change in excitation efficiency caused by spectral drift is highly nonlinear and depends on the working state of the LED, the spectral characteristics of the fluorescence, and the optical path configuration, which is difficult to directly predict by traditional methods.

[0005] Existing LED light source stability control strategies typically use only total optical power as the feedback signal, failing to capture the impact of spectral peak wavelength shifts and bandwidth variations on excitation efficiency. Consequently, even if the light source power itself is stabilized, fluorescence signal instability due to spectral morphology changes persists. Furthermore, existing spectral modeling methods often employ symmetric functions such as Gaussian or Lorentz, which cannot accurately describe the prevalent asymmetric tailing characteristics of LED spectra, leading to insufficient spectral prediction accuracy. Traditional PID control algorithms in LED systems only correct for light intensity fluctuations through feedback control, neglecting spectral changes caused by temperature and current, thus altering excitation efficiency and ultimately resulting in unstable fluorescence signals.

[0006] Based on the aforementioned technical bottlenecks, there is an urgent need for a method to control the stability of LED light sources that compensates for fluorescence emission intensity, thereby stabilizing the fluorescence emission intensity to meet the requirements of quantitative fluorescence imaging for long-term stability and high precision. Summary of the Invention

[0007] The technical problem to be solved by the present invention is to provide a method for controlling LED light source with stable fluorescence emission intensity, so as to solve the problem of fluorescence emission intensity fluctuation caused by large spectral drift and significant peak wavelength shift in existing LED light sources in fluorescence microscopy.

[0008] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0009] A method for controlling the stable fluorescence emission intensity of an LED light source, comprising:

[0010] Step S1: Establish an emission spectrum prediction model for the LED light source to derive the predicted emission spectrum of the LED light source at a given moment based on its junction temperature T and driving current I. This model includes the predicted emission intensity of the LED light source at any emission wavelength λ. ;

[0011] Step S2: When the LED light source is turned on, the fluorescent dye is excited to emit fluorescence in the initial working state. The junction temperature T and driving current I of the LED light source are collected to obtain the initial predicted emission spectrum according to the emission spectrum prediction model, and the target fluorescence emission intensity F is calculated. target :

[0012] ;

[0013] In the formula, The predicted emission intensity is the wavelength λ of the emitted light in the initial predicted emission spectrum. This represents the fluorescence excitation efficiency corresponding to the excitation wavelength λ in the fluorescence excitation spectrum of the fluorescent dye. and These are the upper limit and lower limit values ​​of the excitation wavelength of the fluorescent dye;

[0014] The fluorescence excitation spectrum of the fluorescent dye is a known parameter in the art and can be obtained from standard spectral databases or publicly available technical data from manufacturers. For example, when Rhodamine 123 dye is selected as the fluorescent dye, it has excitation efficiency under excitation light of 300-700 nm. When the wavelength exceeds 700 nm, the excitation efficiency is 0 regardless of the light intensity. Therefore, the target fluorescence emission intensity F is calculated. target The integration range [ , The range is selected as 300-700.

[0015] Step S3: In the working state after the LED light source is turned on, the junction temperature T and driving current I of the LED light source are collected at preset time intervals to obtain the predicted emission spectrum at the t-th collection based on the emission spectrum prediction model, and the fluorescence emission intensity F(t) of the fluorescent dye at the t-th collection is calculated:

[0016] ;

[0017] In the formula, Let be the predicted emission intensity corresponding to the emission wavelength λ in the predicted emission spectrum at the t-th acquisition;

[0018] Step S4: Calculate the fluorescence emission intensity error parameter e(t) at the t-th acquisition in real time:

[0019] ;

[0020] Step S5: During the t-th acquisition, using the fluorescence emission intensity error parameter e(t) as feedback, an incremental PID control algorithm is used to generate the drive current error compensation parameter. And the driving current I of the LED light source is finely controlled to :

[0021] ;

[0022] In the formula, Let e(t), e(t-1), and e(t-2) be the driving current of the LED light source during the t-th acquisition, and e(t-1), e(t-2), respectively, be the fluorescence emission intensity error parameters during the t-th, t-1, and t-2 acquisitions. , , These are the proportional control coefficient, integral control coefficient, and derivative control coefficient, which are dynamically determined by the incremental PID control algorithm at the t-th data acquisition.

[0023] Therefore, this invention uses the fluorescence emission intensity error parameter e(t) calculated by the emission spectrum prediction model based on the LED light source as the feedback for PID control. The emission spectrum prediction model fully considers the changes in the emission spectrum of the LED light source caused by the junction temperature T and the driving current I, and combines the incremental PID control algorithm to generate the driving current error compensation parameter. This allows the driving current I of the LED light source to be adjusted to... This compensates for the changes in fluorescence emission intensity caused by excitation factors such as output power attenuation and spectral drift of the LED light source, ensuring the accuracy and stability of the fluorescence emission intensity emitted by the fluorescent dye excited by the LED light source. Experiments have shown that the actual fluorescence emission intensity can closely approximate the target fluorescence emission intensity F. target The error fluctuation is precisely and stably maintained within 0.1%.

[0024] Preferably, in step S2, after the LED light source is started, the junction temperature T and driving current I of the LED light source are continuously collected five times, and the target fluorescence emission intensity F corresponding to the five collections is calculated according to the formula in step S2. target And from the calculated fluorescence emission intensities F of the five targets target The median is selected as the target fluorescence emission intensity F used in step S4 to calculate the fluorescence emission intensity error parameter e(t) at the t-th acquisition. target To ensure the set target fluorescence emission intensity F target To more closely approximate the steady-state level of the LED light source under actual working conditions, and to avoid interference from abnormal fluorescence emission intensity caused by overshoot or temperature control delay that may occur when the LED light source is turned on; wherein, the sampling time interval between the aforementioned five consecutive samplings of the junction temperature T and driving current I of the LED light source is preferably 1s.

[0025] Preferably, in steps S2 and S3, the target fluorescence emission intensity F is calculated. target The integral range of the emitted light wavelength at fluorescence emission intensity F(t) can be expanded to [ , This ensures full coverage of the excitation wavelength of the fluorescent dye, reducing misjudgments caused by missed wavelengths; among which, and All values ​​are greater than or equal to 0; for example, when Rhodamine 123 dye is selected as the fluorescent dye, expanding the integration range from [300, 700] to [300, 750] can ensure full coverage of the excitation wavelength of Rhodamine 123 dye.

[0026] Preferably, in step S5, the proportional control coefficient is dynamically determined by an incremental PID control algorithm. Integral control coefficient Differential control coefficient The process is as follows:

[0027] Step S5A: Adjust the integral control coefficient Differential control coefficient Set to 0, then gradually increase the proportional control coefficient. Until constant amplitude oscillations occur, the proportional control coefficient is determined. The value of ;

[0028] Step S5B: Maintain the proportional control coefficient determined in step S5A. The value of the integral control coefficient is used for integral control: gradually increase the integral control coefficient. The integral control coefficients are determined only after the steady-state error is eliminated. The value of ;

[0029] Step S5C: Maintain the proportional control coefficient determined in steps S5A and S5B. and integral control coefficient The value of is used for differential control: gradually increase the differential control coefficient. The differential control coefficients are determined until overshoot and oscillation are suppressed. The value of .

[0030] In a preferred embodiment of the present invention, step S1 specifically includes:

[0031] Step S1A: Under the conditions of different junction temperatures T and driving currents I, the corresponding measured emission spectra of the LED light source are obtained by a spectrometer.

[0032] Step S1B: Construct an asymmetric exponential spectral model suitable for the emission characteristics of LED light sources, denoted as the predicted emission spectrum:

[0033] ;

[0034] In the formula, λ is the wavelength of the emitted light from the LED light source. The predicted emitted light intensity of the LED light source is given by the emission wavelength λ. The five parameters to be determined are: peak wavelength λ0, left half peak width parameter a, right half peak width parameter b, peak light intensity parameter c, and asymmetric weighting factor ∂.

[0035] The aforementioned asymmetric exponential spectral model is suitable for the emission characteristics of LED light sources. Its technical concept is as follows: Actual LED spectra typically exhibit asymmetric distribution characteristics, meaning there is a significant difference in decay rates between the left and right sides of the peak. Therefore, the asymmetric exponential spectral model constructed in this invention uses the peak wavelength as a dividing point and employs independent exponential decay functions to describe the low-energy and high-energy sides of the spectrum, respectively. When the emitted light wavelength λ of the LED light source is less than or equal to the peak wavelength λ0, the emitted light intensity is determined by the left-side exponential term controlled by the left half-peak width parameter a. When the emitted light wavelength λ is greater than the peak wavelength λ0, the emitted light intensity is determined by the right-side exponential term controlled by the right half-peak width parameter b. Furthermore, the asymmetric weighting factor ∂ represents the weighting factor of the right-side exponential term relative to the left-side exponential term, and the peak light intensity parameter c is correlated with the measured peak emitted light intensity.

[0036] Step S1C: Substitute the emission wavelength and emission intensity of multiple measured emission spectra under the same junction temperature and current combination conditions in step S1A into the λ and λ of the predicted emission spectrum, respectively. Nonlinear curve fitting was performed to obtain the values ​​of the five undetermined parameters under the junction temperature and current combination conditions;

[0037] Furthermore, based on the values ​​of the five undetermined parameters under multiple different junction temperature and current combinations, the least squares method is used to fit the following: the bivariate quadratic polynomial regression equation of each undetermined parameter (peak wavelength λ0, left half peak width parameter a, right half peak width parameter b, peak light intensity parameter c, asymmetric weighting factor ∂) with respect to the junction temperature and current combination conditions composed of junction temperature T and driving current I, which is denoted as the regression equation of the five undetermined parameters;

[0038] For example, the regression equation for the peak wavelength λ0 is expressed as: The regression equations for the left half-peak width parameter a, the right half-peak width parameter b, the peak light intensity parameter c, the asymmetric weighting factor ∂, and the peak wavelength λ0 are the same, only the fitted coefficients B0, B1, B2, B3, B4, and B5 are different.

[0039] Step S1D: Obtain the emission spectrum prediction model of the LED light source, namely: first, based on the junction temperature T and driving current I of the LED light source, determine the values ​​of peak wavelength λ0, left half-peak width parameter a, right half-peak width parameter b, peak light intensity parameter c, and asymmetric weighting factor ∂ according to the regression equation of the five undetermined parameters; then, substitute the values ​​of the five undetermined parameters into the predicted emission spectrum to determine the predicted emission light intensity of the LED light source at any emission wavelength λ. This enables dynamic prediction of the spectral morphology of LED light sources.

[0040] Therefore, through steps S1A to S1D, this invention employs an asymmetric exponential spectral model suitable for the emission characteristics of LED light sources to construct the predicted emission spectrum of the LED light source. It also combines a bivariate quadratic polynomial regression equation to fit the values ​​of five undetermined parameters of the predicted emission spectrum under different junction temperature-current combinations consisting of junction temperature T and driving current I. This allows the obtained emission spectrum prediction model to accurately describe the spectral morphology changes of the LED light source under different junction temperature-current combinations, including peak wavelength shift, full width at half maximum (FWHM) broadening, and asymmetric characteristics. This is significantly superior to traditional symmetric models. Therefore, this invention improves the prediction accuracy of the emission spectrum of LED light sources based on junction temperature T and driving current I, and further enhances the accuracy and stability of fluorescence emission intensity.

[0041] Preferably, in step S1C, the regression equations for the five undetermined parameters are fitted in the following order:

[0042] First, a bivariate quadratic polynomial regression equation is established for the peak wavelength λ0 with respect to the junction temperature-current combination condition consisting of junction temperature T and driving current I, denoted as the peak wavelength regression equation: λ0 = B0 + B1T + B2I + B3T 2 +B4I 2 +B5T×I; and, substitute the peak wavelengths of multiple measured emission spectra under the same junction temperature and current combination conditions in step S1A into the peak wavelength regression equation, so as to obtain the coefficient values ​​B0, B1, B2, B3, B4, B5 of the peak wavelength regression equation by least squares fitting.

[0043] Then, the value of the peak wavelength λ0 calculated by the peak wavelength regression equation, and the emission wavelength and emission intensity of multiple measured emission spectra under the same junction temperature and current combination conditions in step S1A, are substituted into the predicted emission spectrum's λ0, λ, and λ, respectively. Nonlinear curve fitting was performed to obtain the values ​​of four undetermined parameters under the junction temperature and current combination conditions. The four undetermined parameters are: left half peak width parameter a, right half peak width parameter b, peak light intensity parameter c, and asymmetric weighting factor ∂.

[0044] Finally, based on the values ​​of the four undetermined parameters under multiple different junction temperature and current combinations, the least squares method was used to fit the following: the bivariate quadratic polynomial regression equation of each of the four undetermined parameters with respect to the junction temperature and current combination conditions composed of junction temperature T and driving current I.

[0045] In a preferred embodiment of the present invention: after completing step S1 and before starting step S2, the LED light source control method further includes:

[0046] Step S1-1: Verify the accuracy of the LED light source emission spectrum prediction model established in step S1, including:

[0047] Step S1-1A: Collect the junction temperature T and driving current I of the LED light source under the current operating state to obtain the current predicted emission spectrum based on the emission spectrum prediction model. Simultaneously, measure the actual optical power of the LED light source under the current operating state using a spectrometer. The integral range of the actual optical power measured by the spectrometer should be consistent with the integral range of the relative optical power P described below. , ]same;

[0048] Step S1-1B: Calculate the relative optical power P of the LED light source in the current operating state based on the current predicted emission spectrum.

[0049] ;

[0050] In the formula, The intensity of the predicted emitted light is the wavelength λ in the current predicted emission spectrum.

[0051] Step S1-1C: Compare the relative optical power P with the actual optical power. If the deviation between the two is within the preset optical power error threshold, it means that the LED light source emission spectrum prediction model established in step S1 is accurate and reliable and can be used to execute steps S2 to S5. Otherwise, it means that it is inaccurate and unreliable, and return to re-execute step S1.

[0052] As a preferred embodiment of the present invention, the LED light source control method further includes:

[0053] Step S6: Evaluate the accuracy of the PID control strategy for regulating the driving current I of the LED light source in step S5 in fluorescence imaging where the LED light source excites a fluorescent dye to emit fluorescence, including:

[0054] Step S6A, during the execution of step S5,

[0055] The junction temperature T and driving current I of the LED light source under the current operating state are collected to obtain the current predicted emission spectrum based on the emission spectrum prediction model, and the fluorescence emission intensity change rate K1 and the LED light source excitation power change rate K2 under the current operating state are continuously calculated:

[0056] ;

[0057] ;

[0058] In the formula, F represents the fluorescence emission intensity of the fluorescent dye under the current operating state. The intensity of the predicted emitted light is the wavelength λ in the current predicted emission spectrum. The intensity of the predicted emitted light when the wavelength of the emitted light in the initial predicted emission spectrum is λ;

[0059] Step S6B: Quantitatively evaluate the fluorescence emission intensity F under the current operating state relative to the target fluorescence emission intensity F using the fluorescence emission intensity change rate K1. target Stability: The closer the value of the fluorescence emission intensity change rate K1 is to 1, the more stable the fluorescence emission intensity of the fluorescent dye is.

[0060] Step S6C: Quantitatively evaluate the total output light power fluctuation of the LED light source under the current working state using the LED light source excitation power change rate K2: When the value of the LED light source excitation power change rate K2 is closer to 1, it indicates that the total output light power of the LED light source is more constant.

[0061] Preferably, step S6 further includes:

[0062] Step S6D: Continuously calculate the rate of change of fluorescence excitation efficiency K3 under the current working state:

[0063] ;

[0064] Among them, the fluorescence excitation efficiency change rate K3 reveals the essential reason for the fluorescence signal drift, namely the change in fluorescence excitation efficiency caused by the emission spectrum drift of the LED light source due to temperature or current.

[0065] Step S6E: When the temperature causes a red shift in the spectrum, resulting in a decrease in the fluorescence excitation efficiency change rate K3, the driving current I of the LED light source is increased, thereby increasing the light power of the LED light source and raising the excitation power change rate K2 of the LED light source, so that the fluorescence emission intensity change rate K1 quickly recovers to a stable state, thereby maintaining the stability of the fluorescence emission intensity.

[0066] Conversely, when temperature causes a red shift in the spectrum, leading to an increase in the rate of change of fluorescence excitation efficiency K3, the driving current I of the LED light source is reduced, causing the fluorescence emission intensity to stabilize again.

[0067] The evaluation coefficients are K1 (fluorescence emission intensity change rate), K2 (LED light source excitation power change rate), and K3 (fluorescence excitation efficiency change rate). The technical concept is as follows:

[0068] The fluorescence excitation spectrum of the fluorescent dye and the emission spectrum of the LED light source overlap to some extent. Assuming the peak wavelength λ1 of the fluorescence excitation spectrum and the peak wavelength λ0 of the LED light source emission spectrum, if λ0 > λ1, the initial peak wavelength of the LED light source emission spectrum is located on the longer wavelength side of the fluorescence excitation spectrum. Therefore, the redshift caused by increasing the junction temperature T of the LED light source will further reduce the fluorescence excitation efficiency, resulting in a fluorescence signal attenuation greater than the attenuation of the LED light source output power itself. Conversely, if λ0 < λ1, the initial peak wavelength of the LED light source emission spectrum is located on the shorter wavelength side of the fluorescence excitation spectrum. Therefore, the redshift caused by increasing the junction temperature T of the LED light source may actually enhance the fluorescence excitation efficiency, thereby enhancing the fluorescence signal. Thus, the effects of these enhancement or attenuation effects can be quantitatively analyzed and evaluated using the fluorescence emission intensity change rate K1, the LED light source excitation power change rate K2, and the fluorescence excitation efficiency change rate K3, to reflect the changes in fluorescence emission intensity, light intensity, and fluorescence excitation efficiency rate as the junction temperature T changes.

[0069] Therefore, step S6 of the present invention uses the fluorescence emission intensity change rate K1, the LED light source excitation power change rate K2, and the fluorescence excitation efficiency change rate K3 as evaluation coefficients to accurately characterize the essential perturbation of the fluorescence excitation process: revealing the fundamental reason why traditional optical power feedback cannot suppress fluorescence fluctuations, namely, the change in fluorescence excitation efficiency caused by spectral drift. Thus, it is possible to evaluate the accuracy of the PID control strategy of regulating the driving current I of the LED light source in step S5 in fluorescence imaging when the LED light source excites the fluorescent dye to emit fluorescence.

[0070] As a preferred embodiment of the present invention: the LED light source control method is implemented by a control system consisting of a temperature detection module, a current detection module, a main control module and a memory;

[0071] The temperature detection module is used to acquire the junction temperature T of the LED light source in real time and output the temperature signal to the main control module; the current detection module is used to monitor the driving current I of the LED light source in real time and output the current signal to the main control module; the main control module is used to execute steps S1 to S5, and the memory pre-stores the emission spectrum prediction model of the LED light source in step S1 for the main control module to call.

[0072] The temperature detection module preferably uses a SHT30 digital temperature sensor, mounted on the LED light source substrate with a distance of only 3mm between them, to accurately reflect the real-time junction temperature of the LED light source and ensure measurement accuracy. The current detection module preferably uses a high-precision bidirectional current monitor INA228 to achieve accurate measurement of the drive current I.

[0073] Compared with the prior art, the present invention has the following beneficial effects:

[0074] First, this invention uses the fluorescence emission intensity error parameter e(t) calculated by an LED light source emission spectrum prediction model as feedback for PID control. The emission spectrum prediction model fully considers the changes in the LED light source emission spectrum caused by the junction temperature T and the driving current I, and combines this with an incremental PID control algorithm to generate driving current error compensation parameters. This allows the driving current I of the LED light source to be adjusted to... This compensates for the changes in fluorescence emission intensity caused by excitation factors such as output power attenuation and spectral drift of the LED light source, ensuring the accuracy and stability of the fluorescence emission intensity emitted by the fluorescent dye excited by the LED light source. Experiments have shown that the actual fluorescence emission intensity can closely approximate the target fluorescence emission intensity F. target The error fluctuation is precisely and stably maintained within 0.1%.

[0075] Secondly, through steps S1A to S1D, this invention employs an asymmetric exponential spectral model suitable for the emission characteristics of LED light sources to construct the predicted emission spectrum of the LED light source. It also combines a bivariate quadratic polynomial regression equation to fit the values ​​of five undetermined parameters of the predicted emission spectrum under different junction temperature-current combinations consisting of junction temperature T and driving current I. This allows the obtained emission spectrum prediction model to accurately describe the morphological changes in the emission spectrum of the LED light source under different junction temperature-current combinations, including peak wavelength shift, full width at half maximum (FWHM) broadening, and asymmetric characteristics. This is significantly superior to traditional symmetric models. Therefore, this invention improves the prediction accuracy of the emission spectrum of LED light sources based on junction temperature T and driving current I, further enhancing the accuracy and stability of fluorescence emission intensity.

[0076] Third, step S6 of the present invention uses the fluorescence emission intensity change rate K1, the LED light source excitation power change rate K2, and the fluorescence excitation efficiency change rate K3 as evaluation coefficients, which can accurately characterize the essential perturbation of the fluorescence excitation process: revealing the fundamental reason why traditional optical power feedback cannot suppress fluorescence fluctuations, namely, the change in fluorescence excitation efficiency caused by spectral drift. Thus, it is possible to evaluate the accuracy of the PID control strategy of regulating the driving current I of the LED light source in step S5 in fluorescence imaging when the LED light source excites the fluorescent dye to emit fluorescence.

[0077] Fourth, this invention has the advantages of strong versatility and good real-time performance: the emission spectrum model does not depend on a specific LED model or wavelength, and it is applicable to both single-channel and multi-wavelength LED arrays; the emission spectrum prediction model of the LED light source can be updated in real time to adapt to different microscopic platforms; furthermore, this invention realizes real-time closed-loop control based on the emission spectrum prediction model on an embedded microcontroller, with a response time of milliseconds, which can effectively support the control requirements of high-speed, multi-channel fluorescence imaging, and does not require an external high-performance computing unit. Attached Figure Description

[0078] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments:

[0079] Figure 1 This is a flowchart of the LED light source control method of the present invention;

[0080] Figure 2 The wavelength-emission intensity curve obtained in Experiment 1 of Embodiment 6 of the present invention;

[0081] Figure 3 The junction temperature-emission light intensity curve obtained in Experiment 1 of Embodiment 6 of the present invention;

[0082] Figure 4 This is a schematic diagram of the microscope optical path platform in Experiment 2 as described in Embodiment 6 of the present invention;

[0083] Figure 5 This is a comparison of the fluorescence emission intensity of the LED light source with a center wavelength of 485 nm in Experiment 2;

[0084] Figure 6 This is a comparison of the fluorescence emission intensity of the LED light source with a center wavelength of 505 nm in Experiment 2;

[0085] Figure 7 This is a comparison of the fluorescence emission intensity of the LED light source with a center wavelength of 520 nm in Experiment 2;

[0086] Figure 8 This is a comparison of the fluorescence emission intensity of the LED light source with a center wavelength of 530 nm in Experiment 2. Detailed Implementation

[0087] The present invention will now be described in detail with reference to the embodiments and accompanying drawings to help those skilled in the art better understand the inventive concept of the present invention. However, the scope of protection of the claims of the present invention is not limited to the following embodiments. For those skilled in the art, all other embodiments obtained without creative effort without departing from the inventive concept of the present invention are within the scope of protection of the present invention.

[0088] Example 1

[0089] like Figure 1 As shown, this invention discloses a method for controlling LED light source with stable fluorescence emission intensity, comprising:

[0090] Step S1: Establish an emission spectrum prediction model for the LED light source to derive the predicted emission spectrum of the LED light source at a given moment based on its junction temperature T and driving current I. This model includes the predicted emission intensity of the LED light source at any emission wavelength λ. ;

[0091] Step S2: When the LED light source is turned on, the fluorescent dye is excited to emit fluorescence in the initial working state. The junction temperature T and driving current I of the LED light source are collected to obtain the initial predicted emission spectrum according to the emission spectrum prediction model, and the target fluorescence emission intensity F is calculated. target :

[0092] ;

[0093] In the formula, The predicted emission intensity is the wavelength λ of the emitted light in the initial predicted emission spectrum. This represents the fluorescence excitation efficiency corresponding to the excitation wavelength λ in the fluorescence excitation spectrum of the fluorescent dye. and These are the upper limit and lower limit values ​​of the excitation wavelength of the fluorescent dye;

[0094] The fluorescence excitation spectrum of the fluorescent dye is a known parameter in the art and can be obtained from standard spectral databases or publicly available technical data from manufacturers. For example, when Rhodamine 123 dye is selected as the fluorescent dye, it has excitation efficiency under excitation light of 300-700 nm. When the wavelength exceeds 700 nm, the excitation efficiency is 0 regardless of the light intensity. Therefore, the target fluorescence emission intensity F is calculated. target The integration range [ , The range is selected as 300-700.

[0095] Step S3: In the working state after the LED light source is turned on, the junction temperature T and driving current I of the LED light source are collected at preset time intervals to obtain the predicted emission spectrum at the t-th collection based on the emission spectrum prediction model, and the fluorescence emission intensity F(t) of the fluorescent dye at the t-th collection is calculated:

[0096] ;

[0097] In the formula, Let be the predicted emission intensity corresponding to the emission wavelength λ in the predicted emission spectrum at the t-th acquisition;

[0098] Step S4: Calculate the fluorescence emission intensity error parameter e(t) at the t-th acquisition in real time:

[0099] ;

[0100] Step S5: During the t-th acquisition, using the fluorescence emission intensity error parameter e(t) as feedback, an incremental PID control algorithm is used to generate the drive current error compensation parameter. And the driving current I of the LED light source is finely controlled to :

[0101] ;

[0102] In the formula, Let e(t), e(t-1), and e(t-2) be the driving current of the LED light source during the t-th acquisition, and e(t-1), e(t-2), respectively, be the fluorescence emission intensity error parameters during the t-th, t-1, and t-2 acquisitions. , , These are the proportional control coefficient, integral control coefficient, and derivative control coefficient, which are dynamically determined by the incremental PID control algorithm at the t-th data acquisition.

[0103] Therefore, this invention uses the fluorescence emission intensity error parameter e(t) calculated by the emission spectrum prediction model based on the LED light source as the feedback for PID control. The emission spectrum prediction model fully considers the changes in the emission spectrum of the LED light source caused by the junction temperature T and the driving current I, and combines the incremental PID control algorithm to generate the driving current error compensation parameter. This allows the driving current I of the LED light source to be adjusted to... This compensates for the changes in fluorescence emission intensity caused by excitation factors such as output power attenuation and spectral drift of the LED light source, ensuring the accuracy and stability of the fluorescence emission intensity emitted by the fluorescent dye excited by the LED light source. Experiments have shown that the actual fluorescence emission intensity can closely approximate the target fluorescence emission intensity F. target The error fluctuation is precisely and stably maintained within 0.1%.

[0104] The above is the basic implementation method of this embodiment one, and further optimizations, improvements and limitations can be made based on this basic implementation method:

[0105] Preferably, in step S2, after the LED light source is started, the junction temperature T and driving current I of the LED light source are continuously collected five times, and the target fluorescence emission intensity F corresponding to the five collections is calculated according to the formula in step S2.target And from the calculated fluorescence emission intensities F of the five targets target The median is selected as the target fluorescence emission intensity F used in step S4 to calculate the fluorescence emission intensity error parameter e(t) at the t-th acquisition. target To ensure the set target fluorescence emission intensity F target To more closely approximate the steady-state level of the LED light source under actual working conditions, and to avoid interference from abnormal fluorescence emission intensity caused by overshoot or temperature control delay that may occur when the LED light source is turned on; wherein, the sampling time interval between the aforementioned five consecutive samplings of the junction temperature T and driving current I of the LED light source is preferably 1s.

[0106] Preferably, in steps S2 and S3, the target fluorescence emission intensity F is calculated. target The integral range of the emitted light wavelength at fluorescence emission intensity F(t) can be expanded to [ , This ensures full coverage of the excitation wavelength of the fluorescent dye, reducing misjudgments caused by missed wavelengths; among which, and All values ​​are greater than or equal to 0; for example, when Rhodamine 123 dye is selected as the fluorescent dye, expanding the integration range from [300, 700] to [300, 750] can ensure full coverage of the excitation wavelength of Rhodamine 123 dye.

[0107] Preferably, in step S5, the proportional control coefficient is dynamically determined by an incremental PID control algorithm. Integral control coefficient Differential control coefficient The process is as follows:

[0108] Step S5A: Adjust the integral control coefficient Differential control coefficient Set to 0, then gradually increase the proportional control coefficient. Until constant amplitude oscillations occur, the proportional control coefficient is determined. The value of ;

[0109] Step S5B: Maintain the proportional control coefficient determined in step S5A. The value of the integral control coefficient is used for integral control: gradually increase the integral control coefficient. The integral control coefficients are determined only after the steady-state error is eliminated. The value of ;

[0110] Step S5C: Maintain the proportional control coefficient determined in steps S5A and S5B. and integral control coefficient The value of is used for differential control: gradually increase the differential control coefficient. The differential control coefficients are determined until overshoot and oscillation are suppressed. The value of .

[0111] Example 2

[0112] Based on the above embodiment one, this embodiment two also adopts the following preferred implementation method:

[0113] Step S1 specifically includes:

[0114] Step S1A: Under the conditions of different junction temperatures T and driving currents I, the corresponding measured emission spectra of the LED light source are obtained by a spectrometer.

[0115] Step S1B: Construct an asymmetric exponential spectral model suitable for the emission characteristics of LED light sources, denoted as the predicted emission spectrum:

[0116] ;

[0117] In the formula, λ is the wavelength of the emitted light from the LED light source. The predicted emitted light intensity of the LED light source is given by the emission wavelength λ. The five parameters to be determined are: peak wavelength λ0, left half peak width parameter a, right half peak width parameter b, peak light intensity parameter c, and asymmetric weighting factor ∂.

[0118] The aforementioned asymmetric exponential spectral model is suitable for the emission characteristics of LED light sources. Its technical concept is as follows: Actual LED spectra typically exhibit asymmetric distribution characteristics, meaning there is a significant difference in decay rates between the left and right sides of the peak. Therefore, the asymmetric exponential spectral model constructed in this invention uses the peak wavelength as a dividing point and employs independent exponential decay functions to describe the low-energy and high-energy sides of the spectrum, respectively. When the emitted light wavelength λ of the LED light source is less than or equal to the peak wavelength λ0, the emitted light intensity is determined by the left-side exponential term controlled by the left half-peak width parameter a. When the emitted light wavelength λ is greater than the peak wavelength λ0, the emitted light intensity is determined by the right-side exponential term controlled by the right half-peak width parameter b. Furthermore, the asymmetric weighting factor ∂ represents the weighting factor of the right-side exponential term relative to the left-side exponential term, and the peak light intensity parameter c is correlated with the measured peak emitted light intensity.

[0119] Step S1C: Substitute the emission wavelength and emission intensity of multiple measured emission spectra under the same junction temperature and current combination conditions in step S1A into the λ and λ of the predicted emission spectrum, respectively. Nonlinear curve fitting was performed to obtain the values ​​of the five undetermined parameters under the junction temperature and current combination conditions;

[0120] Furthermore, based on the values ​​of the five undetermined parameters under multiple different junction temperature and current combinations, the least squares method is used to fit the following: the bivariate quadratic polynomial regression equation of each undetermined parameter (peak wavelength λ0, left half peak width parameter a, right half peak width parameter b, peak light intensity parameter c, asymmetric weighting factor ∂) with respect to the junction temperature and current combination conditions composed of junction temperature T and driving current I, which is denoted as the regression equation of the five undetermined parameters;

[0121] For example, the regression equation for the peak wavelength λ0 is expressed as: The regression equations for the left half-peak width parameter a, the right half-peak width parameter b, the peak light intensity parameter c, the asymmetric weighting factor ∂, and the peak wavelength λ0 are the same, only the fitted coefficients B0, B1, B2, B3, B4, and B5 are different.

[0122] Step S1D: Obtain the emission spectrum prediction model of the LED light source, namely: first, based on the junction temperature T and driving current I of the LED light source, determine the values ​​of peak wavelength λ0, left half-peak width parameter a, right half-peak width parameter b, peak light intensity parameter c, and asymmetric weighting factor ∂ according to the regression equation of the five undetermined parameters; then, substitute the values ​​of the five undetermined parameters into the predicted emission spectrum to determine the predicted emission light intensity of the LED light source at any emission wavelength λ. This enables dynamic prediction of the spectral morphology of LED light sources.

[0123] Therefore, through steps S1A to S1D, this invention employs an asymmetric exponential spectral model suitable for the emission characteristics of LED light sources to construct the predicted emission spectrum of the LED light source. It also combines a bivariate quadratic polynomial regression equation to fit the values ​​of five undetermined parameters of the predicted emission spectrum under different junction temperature-current combinations consisting of junction temperature T and driving current I. This allows the obtained emission spectrum prediction model to accurately describe the spectral morphology changes of the LED light source under different junction temperature-current combinations, including peak wavelength shift, full width at half maximum (FWHM) broadening, and asymmetric characteristics. This is significantly superior to traditional symmetric models. Therefore, this invention improves the prediction accuracy of the emission spectrum of LED light sources based on junction temperature T and driving current I, and further enhances the accuracy and stability of fluorescence emission intensity.

[0124] The above is the basic implementation method of this embodiment two, and further optimizations, improvements and limitations can be made based on this basic implementation method:

[0125] Preferably, in step S1C, the regression equations for the five undetermined parameters are fitted in the following order:

[0126] First, a bivariate quadratic polynomial regression equation is established for the peak wavelength λ0 with respect to the junction temperature-current combination condition consisting of junction temperature T and driving current I, denoted as the peak wavelength regression equation: λ0 = B0 + B1T + B2I + B3T 2 +B4I 2 +B5T×I; and, substitute the peak wavelengths of multiple measured emission spectra under the same junction temperature and current combination conditions in step S1A into the peak wavelength regression equation, so as to obtain the coefficient values ​​B0, B1, B2, B3, B4, B5 of the peak wavelength regression equation by least squares fitting.

[0127] Then, the value of the peak wavelength λ0 calculated by the peak wavelength regression equation, and the emission wavelength and emission intensity of multiple measured emission spectra under the same junction temperature and current combination conditions in step S1A, are substituted into the predicted emission spectrum's λ0, λ, and λ, respectively. Nonlinear curve fitting was performed to obtain the values ​​of four undetermined parameters under the junction temperature and current combination conditions. The four undetermined parameters are: left half peak width parameter a, right half peak width parameter b, peak light intensity parameter c, and asymmetric weighting factor ∂.

[0128] Finally, based on the values ​​of the four undetermined parameters under multiple different junction temperature and current combinations, the least squares method was used to fit the following: the bivariate quadratic polynomial regression equation of each of the four undetermined parameters with respect to the junction temperature and current combination conditions composed of junction temperature T and driving current I.

[0129] Example 3

[0130] Based on the above embodiment one or embodiment two, this embodiment three also adopts the following preferred implementation method:

[0131] After completing step S1 and before starting step S2, the LED light source control method further includes:

[0132] Step S1-1: Verify the accuracy of the LED light source emission spectrum prediction model established in step S1, including:

[0133] Step S1-1A: Collect the junction temperature T and driving current I of the LED light source under the current operating state to obtain the current predicted emission spectrum based on the emission spectrum prediction model. Simultaneously, measure the actual optical power of the LED light source under the current operating state using a spectrometer. The integral range of the actual optical power measured by the spectrometer should be consistent with the integral range of the relative optical power P described below. , ]same;

[0134] Step S1-1B: Calculate the relative optical power P of the LED light source in the current operating state based on the current predicted emission spectrum.

[0135] ;

[0136] In the formula, The intensity of the predicted emitted light is the wavelength λ in the current predicted emission spectrum.

[0137] Step S1-1C: Compare the relative optical power P with the actual optical power. If the deviation between the two is within the preset optical power error threshold, it means that the LED light source emission spectrum prediction model established in step S1 is accurate and reliable and can be used to execute steps S2 to S5. Otherwise, it means that it is inaccurate and unreliable, and return to re-execute step S1.

[0138] Example 4

[0139] Based on any one of the above embodiments one to three, this embodiment four further adopts the following preferred implementation method:

[0140] The LED light source control method further includes:

[0141] Step S6: Evaluate the accuracy of the PID control strategy for regulating the driving current I of the LED light source in step S5 in fluorescence imaging where the LED light source excites a fluorescent dye to emit fluorescence, including:

[0142] Step S6A, during the execution of step S5,

[0143] The junction temperature T and driving current I of the LED light source under the current operating state are collected to obtain the current predicted emission spectrum based on the emission spectrum prediction model, and the fluorescence emission intensity change rate K1 and the LED light source excitation power change rate K2 under the current operating state are continuously calculated:

[0144] ;

[0145] ;

[0146] In the formula, F represents the fluorescence emission intensity of the fluorescent dye under the current operating state. The intensity of the predicted emitted light is the wavelength λ in the current predicted emission spectrum. The intensity of the predicted emitted light when the wavelength of the emitted light in the initial predicted emission spectrum is λ;

[0147] Step S6B: Quantitatively evaluate the fluorescence emission intensity F under the current operating state relative to the target fluorescence emission intensity F using the fluorescence emission intensity change rate K1. targetStability: The closer the value of the fluorescence emission intensity change rate K1 is to 1, the more stable the fluorescence emission intensity of the fluorescent dye is.

[0148] Step S6C: Quantitatively evaluate the total output light power fluctuation of the LED light source under the current working state using the LED light source excitation power change rate K2: When the value of the LED light source excitation power change rate K2 is closer to 1, it indicates that the total output light power of the LED light source is more constant.

[0149] The above is the basic implementation method of this embodiment four, and further optimizations, improvements and limitations can be made based on this basic implementation method:

[0150] Preferably, step S6 further includes:

[0151] Step S6D: Continuously calculate the rate of change of fluorescence excitation efficiency K3 under the current working state:

[0152] ;

[0153] Among them, the fluorescence excitation efficiency change rate K3 reveals the essential reason for the fluorescence signal drift, namely the change in fluorescence excitation efficiency caused by the emission spectrum drift of the LED light source due to temperature or current.

[0154] Step S6E: When the temperature causes a red shift in the spectrum, resulting in a decrease in the fluorescence excitation efficiency change rate K3, the driving current I of the LED light source is increased, thereby increasing the light power of the LED light source and raising the excitation power change rate K2 of the LED light source, so that the fluorescence emission intensity change rate K1 quickly recovers to a stable state, thereby maintaining the stability of the fluorescence emission intensity.

[0155] Conversely, when temperature causes a red shift in the spectrum, leading to an increase in the rate of change of fluorescence excitation efficiency K3, the driving current I of the LED light source is reduced, causing the fluorescence emission intensity to stabilize again.

[0156] The evaluation coefficients are K1 (fluorescence emission intensity change rate), K2 (LED light source excitation power change rate), and K3 (fluorescence excitation efficiency change rate). The technical concept is as follows:

[0157] The fluorescence excitation spectrum of the fluorescent dye and the emission spectrum of the LED light source overlap to some extent. Assuming the peak wavelength λ1 of the fluorescence excitation spectrum and the peak wavelength λ0 of the LED light source emission spectrum, if λ0 > λ1, the initial peak wavelength of the LED light source emission spectrum is located on the longer wavelength side of the fluorescence excitation spectrum. Therefore, the redshift caused by increasing the junction temperature T of the LED light source will further reduce the fluorescence excitation efficiency, resulting in a fluorescence signal attenuation greater than the attenuation of the LED light source output power itself. Conversely, if λ0 < λ1, the initial peak wavelength of the LED light source emission spectrum is located on the shorter wavelength side of the fluorescence excitation spectrum. Therefore, the redshift caused by increasing the junction temperature T of the LED light source may actually enhance the fluorescence excitation efficiency, thereby enhancing the fluorescence signal. Thus, the effects of these enhancement or attenuation effects can be quantitatively analyzed and evaluated using the fluorescence emission intensity change rate K1, the LED light source excitation power change rate K2, and the fluorescence excitation efficiency change rate K3, to reflect the changes in fluorescence emission intensity, light intensity, and fluorescence excitation efficiency rate as the junction temperature T changes.

[0158] Therefore, step S6 of the present invention uses the fluorescence emission intensity change rate K1, the LED light source excitation power change rate K2, and the fluorescence excitation efficiency change rate K3 as evaluation coefficients to accurately characterize the essential perturbation of the fluorescence excitation process: revealing the fundamental reason why traditional optical power feedback cannot suppress fluorescence fluctuations, namely, the change in fluorescence excitation efficiency caused by spectral drift. Thus, it is possible to evaluate the accuracy of the PID control strategy of regulating the driving current I of the LED light source in step S5 in fluorescence imaging when the LED light source excites the fluorescent dye to emit fluorescence.

[0159] Example 5

[0160] Based on any one of the above embodiments one to four, this embodiment five also adopts the following preferred implementation method:

[0161] The LED light source control method is implemented by a control system consisting of a temperature detection module, a current detection module, a main control module, and a memory.

[0162] The temperature detection module is used to acquire the junction temperature T of the LED light source in real time and output the temperature signal to the main control module; the current detection module is used to monitor the driving current I of the LED light source in real time and output the current signal to the main control module; the main control module is used to execute steps S1 to S5, and the memory pre-stores the emission spectrum prediction model of the LED light source in step S1 for the main control module to call.

[0163] The temperature detection module preferably uses a SHT30 digital temperature sensor, mounted on the LED light source substrate with a distance of only 3mm between them, to accurately reflect the real-time junction temperature of the LED light source and ensure measurement accuracy. The current detection module preferably uses a high-precision bidirectional current monitor INA228 to achieve accurate measurement of the drive current I.

[0164] Example 6

[0165] The purpose of this sixth embodiment is to design and implement a set of test experiments to verify the accuracy of the emission spectrum model and the changes in fluorescence emission intensity, so as to verify the practicality and stability of the present invention.

[0166] The steps of Experiment 1 are as follows:

[0167] To verify the accuracy and effectiveness of the LED light source emission spectrum prediction model described in this invention, this experiment provides a calibration and verification process based on the spectral model. This process is implemented based on experimental equipment, including: a customized multi-wavelength LED light source (center wavelengths of 485 nm, 505 nm, 520 nm, and 530 nm, respectively) as the LED light source, a high-resolution spectrometer (Ocean Optics QE65Pro, wavelength range 200–1100 nm, resolution 0.8 nm), and a 32-bit microcontroller (STM32F103RCT6, STMicroelectronics) as the main control module.

[0168] Following step S1, which includes steps S1A to S1D, experiments were conducted on multi-wavelength LED light sources, specifically four LED light sources with center wavelengths (485 nm, 505 nm, 520 nm, and 530 nm, respectively), to obtain... Figure 2 The experimental results of the wavelength-emitted light intensity curves are shown. The four colors represent the four LED light sources corresponding to the four center wavelengths. The straight line represents the predicted result calculated in step S1 of this invention, and the dotted line represents the actual result measured by the spectrometer. It can be seen that the four LED light sources (center wavelengths of 485 nm, 505 nm, 520 nm, and 530 nm, respectively) have similar temperature and current characteristics. The asymmetric exponential spectral model proposed in this invention achieves high-precision fitting of all measured spectra, with an evaluation coefficient R... 2 The coefficient R0.99 confirms that the emission spectrum prediction model of this invention effectively describes the asymmetric spectral distribution of LED light sources. Furthermore, the regression model for the five undetermined parameters of the predicted emission spectrum of this invention has evaluation coefficients R0.99. 2 All are above 0.95, representing the evaluation coefficient R based on the nonlinear fitting of the spectral model and the actual spectrum. 2All values ​​are above 0.99, indicating that junction temperature and current can effectively characterize the evolution of spectral parameters.

[0169] Furthermore, following steps S1-1, including S1-1A to S1-1C, experiments were conducted on multi-wavelength LED light sources, specifically LED light sources with four center wavelengths (485 nm, 505 nm, 520 nm, and 530 nm, respectively), and the results were obtained. Figure 3 The experimental results of the junction temperature-emission light intensity curves are shown. The four colors represent the four LED light sources corresponding to the four center wavelengths. The straight line represents the prediction result calculated by step S1-1B of the present invention, and the dotted line represents the actual result measured by the spectrometer. This verifies that the LED light source emission spectrum prediction model established in step S1 is accurate.

[0170] The steps of Experiment 2 are as follows:

[0171] To verify the changes in fluorescence emission intensity and the effectiveness of the LED light source emission spectrum model in achieving stable control of fluorescence emission intensity in a practical imaging system, this second experiment is based on... Figure 4 The microscope optical path platform shown includes: an achromatic objective lens 1 (60×1.42NA oil, Olympus) for a fluorescence microscope, a power meter 2 (PM400, ThorLabs), a 50:50 beam splitter 3 (LABTEK), a CMOS camera 4 (ORCA-Flash 4.0 V3, Hamamatsu), and a four-wavelength LED light source 5 (center wavelengths of 485nm, 505nm, 520nm, and 530nm, respectively), an excitation filter 6, a dichroic mirror 7, and a mirror 9. The LED excitation module is integrated into a square fluorescence microscope based on the Olympus IX73 platform. Furthermore, using Rhodamine 123 dye as the fluorescent sample 8, the effectiveness of the control strategy of this invention is verified through multi-band excitation.

[0172] Following steps S2 to S6 above, experiments were conducted on four wavelength LED light sources 5, that is, four LED light sources with four center wavelengths (485 nm, 505 nm, 520 nm, and 530 nm, respectively), and the results were obtained. Figures 5 to 8 The graph showing the comparison of fluorescence emission intensity of four LED light sources represents a complete record of the change in fluorescence emission intensity over time. The solid line represents the curve of change in optical power calculated by this invention, the dotted line represents the grayscale value of the image captured by CMOS camera 4, which is used to reflect the change in fluorescence emission intensity, and the three color curves represent the curves of change in optical power and fluorescence emission intensity under the condition of fluorescence emission intensity feedback adjustment: gray represents no feedback adjustment, blue represents only optical power feedback adjustment, and red represents the curves of change in optical power and fluorescence emission intensity under the condition of fluorescence emission intensity feedback adjustment.

[0173] By analyzing the trends in fluorescence emission intensity and optical power fluctuations, the stability improvement effect of the closed-loop control PID method of this invention under different wavelengths and operating conditions is quantitatively evaluated. Figures 5 to 8 The fluorescence experimental results at the four wavelengths shown all demonstrate that this invention, by regulating the output state of the LED light source based on a spectral model, reduces fluorescence fluctuations from 2.18%-9.41% to 0.44%-1.25%, enabling the system to maintain a high degree of stability in the fluorescence signal even under spectral shifts and optical power fluctuations. Compared with traditional stability control methods that only address optical power feedback, this invention also comprehensively considers the essential factors of spectral distribution and excitation efficiency changes, effectively eliminating the influence of optical power attenuation and wavelength shifts on fluorescence excitation, and significantly improving the stability and repeatability of the fluorescence signal.

[0174] Because the control strategy of this invention no longer relies on the stability of the optical power itself, but instead predicts and corrects the excitation efficiency based on a spectral model, the fluorescence emission intensity can still be maintained at a steady-state level even when the LED light source experiences spectral shifts due to temperature disturbances and current fluctuations. This avoids the problem of bias accumulation in long-term experiments using traditional methods. This invention significantly improves the reliability of fluorescence experimental data and can be widely applied in fluorescence microscopy imaging systems and multi-wavelength LED light source devices.

[0175] This invention is not limited to the specific embodiments described above. Based on the above content and in accordance with common technical knowledge and conventional methods in the field, without departing from the basic technical concept of this invention, this invention can also make other equivalent modifications, substitutions or alterations, all of which fall within the protection scope of this invention.

Claims

1. A method for controlling the stable fluorescence emission intensity of an LED light source, characterized in that, include: Step S1: Establish an emission spectrum prediction model for the LED light source to derive the predicted emission spectrum of the LED light source at a given moment based on its junction temperature T and driving current I. This model includes the predicted emission intensity of the LED light source at any emission wavelength λ. ; Step S2: When the LED light source is turned on, the fluorescent dye is excited to emit fluorescence in the initial working state. The junction temperature T and driving current I of the LED light source are collected to obtain the initial predicted emission spectrum according to the emission spectrum prediction model, and the target fluorescence emission intensity F is calculated. target ; Step S3: In the working state after the LED light source is turned on, the junction temperature T and driving current I of the LED light source are collected at preset time intervals to obtain the predicted emission spectrum at the tth collection according to the emission spectrum prediction model, and the fluorescence emission intensity F(t) of the fluorescent dye at the tth collection is calculated. Step S4: Calculate the fluorescence emission intensity error parameter e(t) at the t-th acquisition in real time; Step S5: During the t-th acquisition, using the fluorescence emission intensity error parameter e(t) as feedback, an incremental PID control algorithm is used to generate the drive current error compensation parameter. And the driving current I of the LED light source is adjusted to .

2. The LED light source control method for stabilizing fluorescence emission intensity according to claim 1, characterized in that: In step S2, after the LED light source is started, the junction temperature T and driving current I of the LED light source are continuously collected five times, and the target fluorescence emission intensity F corresponding to the five collections is calculated according to the formula in step S2. target And from the calculated fluorescence emission intensities F of the five targets target The median is selected as the target fluorescence emission intensity F used in step S4 to calculate the fluorescence emission intensity error parameter e(t) at the t-th acquisition. target .

3. The LED light source control method for stabilizing fluorescence emission intensity according to claim 1, characterized in that: In steps S2 and S3, the target fluorescence emission intensity F is calculated. target The integral range of the emitted light wavelength at fluorescence emission intensity F(t) is expanded to [ , ].

4. The LED light source control method for stabilizing fluorescence emission intensity according to claim 1, characterized in that: In step S5, the proportional control coefficient is dynamically determined by the incremental PID control algorithm. Integral control coefficient Differential control coefficient The process is as follows: Step S5A: Adjust the integral control coefficient Differential control coefficient Set to 0, then gradually increase the proportional control coefficient. Until constant amplitude oscillations occur, the proportional control coefficient is determined. The value of ; Step S5B: Maintain the proportional control coefficient determined in step S5A. The value of the integral control coefficient is used for integral control: gradually increase the integral control coefficient. The integral control coefficients are determined only after the steady-state error is eliminated. The value of ; Step S5C: Maintain the proportional control coefficient determined in steps S5A and S5B. and integral control coefficient The value of is used for differential control: gradually increase the differential control coefficient. The differential control coefficients are determined until overshoot and oscillation are suppressed. The value of .

5. The LED light source control method for stabilizing fluorescence emission intensity according to any one of claims 1 to 4, characterized in that: Step S1 specifically includes: Step S1A: Under the conditions of different junction temperatures T and driving currents I, the corresponding measured emission spectra of the LED light source are obtained by a spectrometer. Step S1B: Construct an asymmetric exponential spectral model suitable for the emission characteristics of LED light sources, denoted as the predicted emission spectrum: ； In the formula, λ is the wavelength of the emitted light from the LED light source. The predicted emitted light intensity of the LED light source is given by the emission wavelength λ. The five parameters to be determined are: peak wavelength λ0, left half peak width parameter a, right half peak width parameter b, peak light intensity parameter c, and asymmetric weighting factor ∂. Step S1C: Substitute the emission wavelength and emission intensity of multiple measured emission spectra under the same junction temperature and current combination conditions in step S1A into the λ and λ of the predicted emission spectrum, respectively. Nonlinear curve fitting was performed to obtain the values ​​of the five undetermined parameters under the junction temperature and current combination conditions; Furthermore, based on the values ​​of the five undetermined parameters under multiple different junction temperature and current combinations, the least squares method is used to fit the following: the bivariate quadratic polynomial regression equation of each undetermined parameter with respect to the junction temperature and current combination conditions composed of junction temperature T and driving current I, which is denoted as the regression equation of the five undetermined parameters. Step S1D: Obtain the emission spectrum prediction model of the LED light source, namely: first, based on the junction temperature T and driving current I of the LED light source, determine the values ​​of peak wavelength λ0, left half-peak width parameter a, right half-peak width parameter b, peak light intensity parameter c, and asymmetric weighting factor ∂ according to the regression equation of the five undetermined parameters; then, substitute the values ​​of the five undetermined parameters into the predicted emission spectrum to determine the predicted emission light intensity of the LED light source at any emission wavelength λ. .

6. The LED light source control method for stabilizing fluorescence emission intensity according to claim 5, characterized in that: In step S1C, the regression equations for the five undetermined parameters are fitted in the following order: First, a bivariate quadratic polynomial regression equation is established for the peak wavelength λ0 with respect to the junction temperature-current combination condition consisting of junction temperature T and driving current I, denoted as the peak wavelength regression equation: λ0 = B0 + B1T + B2I + B3T 2 +B4I 2 +B5T×I; and, substitute the peak wavelengths of multiple measured emission spectra under the same junction temperature and current combination conditions in step S1A into the peak wavelength regression equation, so as to obtain the coefficient values ​​B0, B1, B2, B3, B4, B5 of the peak wavelength regression equation by least squares fitting. Then, the value of the peak wavelength λ0 calculated by the peak wavelength regression equation, and the emission wavelength and emission intensity of multiple measured emission spectra under the same junction temperature and current combination conditions in step S1A, are substituted into the predicted emission spectrum's λ0, λ, and λ, respectively. Nonlinear curve fitting was performed to obtain the values ​​of four undetermined parameters under the junction temperature and current combination conditions. The four undetermined parameters are: left half peak width parameter a, right half peak width parameter b, peak light intensity parameter c, and asymmetric weighting factor ∂. Finally, based on the values ​​of the four undetermined parameters under multiple different junction temperature and current combinations, the least squares method was used to fit the following: the bivariate quadratic polynomial regression equation of each of the four undetermined parameters with respect to the junction temperature and current combination conditions composed of junction temperature T and driving current I.

7. The LED light source control method for stabilizing fluorescence emission intensity according to any one of claims 1 to 4, characterized in that: After completing step S1 and before starting step S2, the LED light source control method further includes: Step S1-1: Verify the accuracy of the LED light source emission spectrum prediction model established in step S1, including: Step S1-1A: Collect the junction temperature T and driving current I of the LED light source under the current working state, so as to obtain the current predicted emission spectrum according to the emission spectrum prediction model. At the same time, use a spectrometer to measure the actual optical power of the LED light source under the current working state. Step S1-1B: Calculate the relative optical power P of the LED light source in the current operating state based on the current predicted emission spectrum. ； In the formula, The intensity of the predicted emitted light is the wavelength λ in the current predicted emission spectrum. Step S1-1C: Compare the relative optical power P with the actual optical power. If the deviation between the two is within the preset optical power error threshold, it means that the LED light source emission spectrum prediction model established in step S1 is accurate and reliable and can be used to execute steps S2 to S5. Otherwise, it means that it is inaccurate and unreliable, and return to re-execute step S1.

8. The LED light source control method for stabilizing fluorescence emission intensity according to any one of claims 1 to 4, characterized in that: The LED light source control method further includes: Step S6: Evaluate the accuracy of the PID control strategy for regulating the driving current I of the LED light source in step S5 in fluorescence imaging where the LED light source excites a fluorescent dye to emit fluorescence, including: Step S6A, during the execution of step S5, The junction temperature T and driving current I of the LED light source under the current working state are collected to obtain the current predicted emission spectrum based on the emission spectrum prediction model, and to continuously calculate the fluorescence emission intensity change rate K1 and the LED light source excitation power change rate K2 under the current working state. Step S6B: Quantitatively evaluate the fluorescence emission intensity F under the current operating state relative to the target fluorescence emission intensity F using the fluorescence emission intensity change rate K1. target Stability: The closer the value of the fluorescence emission intensity change rate K1 is to 1, the more stable the fluorescence emission intensity of the fluorescent dye is. Step S6C: Quantitatively evaluate the total output light power fluctuation of the LED light source under the current working state using the LED light source excitation power change rate K2: When the value of the LED light source excitation power change rate K2 is closer to 1, it indicates that the total output light power of the LED light source is more constant.

9. The LED light source control method for stabilizing fluorescence emission intensity according to claim 8, characterized in that: Step S6 further includes: Step S6D: Continuously calculate the rate of change of fluorescence excitation efficiency K3 under the current working state: ； Step S6E: When the fluorescence excitation efficiency change rate K3 decreases, the driving current I of the LED light source is increased. Conversely, when the fluorescence excitation efficiency change rate K3 increases, the driving current I of the LED light source decreases.

10. The LED light source control method for stabilizing fluorescence emission intensity according to any one of claims 1 to 4, characterized in that: The LED light source control method is implemented by a control system consisting of a temperature detection module, a current detection module, a main control module, and a memory. The temperature detection module is used to acquire the junction temperature T of the LED light source in real time and output the temperature signal to the main control module; the current detection module is used to monitor the driving current I of the LED light source in real time and output the current signal to the main control module; the main control module is used to execute steps S1 to S5, and the memory pre-stores the emission spectrum prediction model of the LED light source in step S1 for the main control module to call.

Images