A LED display screen brightness monitoring method and system

Abstract

This invention relates to the field of display control, specifically to a method and system for monitoring the brightness of an LED display screen. The method involves acquiring the real-time junction temperature, cumulative driving pulse count, and pulse width distribution parameters of each LED pixel, inputting these parameters into a nonlinear coupling prediction model of light decay-junction temperature-driving history to output a predicted brightness decay coefficient. Based on this predicted brightness decay coefficient, the grayscale and current mapping relationship is reconstructed. When the decay coefficient is below a first set threshold, the constant current source reference current is kept constant while the pulse width modulation duty cycle is increased to compensate for brightness. When the decay coefficient exceeds the first set threshold, the constant current source reference current value is increased and the pulse width modulation duty cycle is adjusted back proportionally. This eliminates brightness unevenness caused by pixel-level nonlinear light decay differences, overcomes the bottleneck of a single duty cycle compensation limit, and ensures consistent compensation across high and low grayscale ranges.

Description

Technical Field

[0001] This invention relates to the field of display control, specifically to a method and system for monitoring the brightness of an LED display screen. Background Technology

[0002] During long-term operation, the brightness of the light-emitting devices in an LED display will decrease over time. Existing conventional brightness compensation schemes typically employ a lookup table-based compensation method for overall brightness attenuation based on operating time. This method involves pre-storing a table in the controller that corresponds to the display's operating time and the overall brightness attenuation coefficient. When the display reaches a specific time point, a uniform brightness attenuation coefficient is obtained by looking up the table. Based on this uniform attenuation coefficient, a pulse width modulation duty cycle is increased by the same proportion for all the emitting pixels on the screen to compensate for the brightness loss.

[0003] The differences in real-time junction temperature and historical driving pulse width distribution among pixels in a light-emitting diode display (LED) screen cause non-linear characteristics in pixel-level light decay. Compensation methods based on runtime and unified lookup tables can only perform post-event and globally uniform adjustments, and cannot perform pre-prediction and differentiated compensation for the non-linear light decay differences of individual pixels, resulting in uneven brightness after long-term operation of the display panel. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for monitoring the brightness of an LED display screen, which can effectively solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] A method for monitoring the brightness of an LED display screen includes: acquiring the real-time junction temperature of each LED pixel in the LED display screen through a junction temperature sensing network;

[0007] The cumulative number of driving pulses and pulse width distribution parameters of each LED pixel are recorded by driving a historical statistical array.

[0008] The real-time junction temperature, the cumulative number of driving pulses, and the pulse width distribution parameters are input into the nonlinear coupling prediction model of light decay-junction temperature-driving history, and the expected brightness decay coefficient of each LED pixel is output.

[0009] Based on the expected brightness attenuation coefficient, the mapping relationship between grayscale and current is reconstructed. When the expected brightness attenuation coefficient is lower than the first set threshold, the constant current source reference current is kept unchanged, and the brightness is compensated by increasing the pulse width modulation duty cycle.

[0010] When the expected brightness attenuation coefficient exceeds the first set threshold, the constant current source reference current value is increased and the pulse width modulation duty cycle is adjusted back proportionally.

[0011] Preferably, obtaining the real-time junction temperature of each LED pixel in the LED display screen through the junction temperature sensing network includes: during the line scan blanking period, applying a reverse bias test pulse of a specific frequency to the line line where the LED pixel is located, and keeping the duration of the reverse bias test pulse less than the total duration of the line blanking period.

[0012] The transient voltage and transient current changes of the line under the reverse bias test pulse are collected.

[0013] The dynamic impedance parameters of the LED pixel are calculated based on the transient voltage change value and the transient current change value;

[0014] The dynamic impedance parameter is interpolated and matched with the pre-calibrated impedance-temperature correlation curve, and the temperature value corresponding to the dynamic impedance parameter is extracted as the real-time junction temperature of each LED pixel.

[0015] Preferably, the step of recording the cumulative number of driving pulses and pulse width distribution parameters of each LED pixel through the driving history statistical array includes: for each LED pixel, counting the effective pulse counts of the LED pixel being turned on in each frame scan cycle;

[0016] The effective pulse count is added to the cumulative drive pulse count;

[0017] Simultaneously, the grayscale range of the positive driving pulse width corresponding to the LED pixel in each frame is identified, and corresponding weighting coefficients are set for different grayscale ranges;

[0018] The positive driving pulse width is multiplied by the corresponding weighting coefficient and then accumulated to generate a weighted pulse width cumulative value. The weighted pulse width cumulative value and the cumulative driving pulse array are combined to form the pulse width distribution parameter.

[0019] Preferably, the real-time junction temperature, the cumulative number of driving pulses, and the pulse width distribution parameters are input into the nonlinear coupling prediction model of light decay-junction temperature-driving history, and the expected brightness decay coefficient of each LED pixel is output by: using the real-time junction temperature as a thermal acceleration factor, and using the cumulative number of driving pulses and the pulse width distribution parameters as a cumulative degradation factor.

[0020] In the nonlinear coupling prediction model of optical decay-junction temperature-drive history, an exponential cross-product term is constructed between the thermal acceleration factor and the cumulative degradation factor.

[0021] The value of the exponential cross product term is calculated and normalized to obtain the expected brightness decay coefficient of each LED pixel. The expected brightness decay coefficient reflects the decay ratio of the current pixel brightness relative to the initial brightness.

[0022] Preferably, when the expected brightness attenuation coefficient is lower than the first set threshold, maintaining the constant current source reference current unchanged and compensating for brightness by increasing the pulse width modulation duty cycle includes: when it is determined that the expected brightness attenuation coefficient is lower than the first set threshold, locking the reference voltage of the constant current source reference current to the initial calibration value;

[0023] Calculate the required brightness compensation gain value based on the expected brightness attenuation coefficient;

[0024] The target grayscale data is determined based on the product of the brightness compensation gain value and the original grayscale data.

[0025] The target grayscale data is remapped to the corresponding pulse width modulation duty cycle so that the actual output pulse width modulation duty cycle is greater than the duty cycle corresponding to the original grayscale data, thus completing the brightness compensation.

[0026] Preferably, when the expected brightness attenuation coefficient exceeds the first set threshold, increasing the constant current source reference current reference value and proportionally reducing the pulse width modulation duty cycle includes: when it is determined that the expected brightness attenuation coefficient exceeds the first set threshold, determining the increase step size of the constant current source reference current according to the attenuation range where the expected brightness attenuation coefficient is located, and increasing the reference voltage of the constant current source reference current according to the increase step size;

[0027] Meanwhile, the duty cycle callback coefficient is determined based on the ratio of the enhanced constant current source reference current to the initial calibration value.

[0028] The pulse width modulation duty cycle corresponding to the original grayscale data is multiplied by the duty cycle callback coefficient to generate the updated pulse width modulation duty cycle.

[0029] Preferably, the process of collecting the transient voltage change value and transient current change value of the horizontal line under the action of the reverse bias test pulse further includes: collecting the initial parasitic capacitance voltage and leakage current of the horizontal line before applying the reverse bias test pulse of the specific frequency;

[0030] The line parasitic effect offset is calculated based on the initial parasitic capacitance voltage and leakage current.

[0031] When calculating the dynamic impedance parameters, the line parasitic effect offset is subtracted from the transient voltage change value and transient current change value to generate corrected transient voltage change value and transient current change value, and the dynamic impedance parameters are calculated using the corrected values.

[0032] Preferably, setting corresponding weighting coefficients for different grayscale ranges further includes: obtaining the cumulative working time of the LED pixel during continuous operation;

[0033] Based on the cumulative working time, a preset weighted coefficient attenuation table is searched to obtain the attenuation correction factor corresponding to the cumulative working time;

[0034] The weighting coefficients initially set are multiplied by the attenuation correction factor to generate dynamically updated weighting coefficients;

[0035] As the cumulative working time increases, the attenuation correction factor decreases monotonically, causing the weight of the weighting coefficient corresponding to the high grayscale interval in the total pulse width distribution parameter to decrease over time. The dynamically updated weighting coefficient participates in the calculation of the weighted pulse width cumulative value in the next frame.

[0036] Preferably, increasing the reference voltage of the constant current source reference current according to the increase step size further includes: after increasing the reference voltage of the constant current source reference current in the row where the target LED pixel is located, calculating the reference voltage difference between the target LED pixel and the adjacent LED pixel;

[0037] If the reference voltage difference exceeds the set voltage step threshold, the lifting step size of the target LED pixel is divided into levels, and the single lifting step size is replaced with a multi-level progressive lifting step size.

[0038] Within multiple consecutive frame cycles, the constant current source reference value of the target LED pixel is gradually increased according to the multi-level progressive increase step size, while the duty cycle callback coefficient is adjusted synchronously step by step.

[0039] An LED display screen brightness monitoring system includes: a junction temperature sensing network for acquiring the real-time junction temperature of each LED pixel in the LED display screen;

[0040] A historical statistics array is used to record the cumulative number of driving pulses and pulse width distribution parameters of each LED pixel;

[0041] The light decay prediction calculator is used to input the real-time junction temperature, the cumulative number of driving pulses and the pulse width distribution parameters into the light decay-junction temperature-driving history nonlinear coupling prediction model, and output the expected brightness decay coefficient of each LED pixel.

[0042] A hybrid drive logic controller is used to reconstruct the mapping relationship between grayscale and current based on the expected brightness attenuation coefficient. When the expected brightness attenuation coefficient is lower than a first set threshold, the constant current source reference current is kept unchanged, and the brightness is compensated by increasing the pulse width modulation duty cycle.

[0043] When the expected brightness attenuation coefficient exceeds the first set threshold, the constant current source reference current value is increased and the pulse width modulation duty cycle is adjusted back proportionally.

[0044] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0045] 1. This invention inputs the real-time junction temperature, cumulative driving pulse count, and pulse width distribution parameters of each pixel into a nonlinear coupling prediction model of light decay-junction temperature-driving history, outputs the expected brightness decay coefficient, and reconstructs the mapping relationship between grayscale and current based on the expected brightness decay coefficient. When the expected brightness decay coefficient is lower than a first set threshold, the constant current source reference current is kept constant, and the pulse width modulation duty cycle is increased to compensate for brightness. When the expected brightness decay coefficient exceeds the first set threshold, the reference value of the constant current source reference current is increased and the pulse width modulation duty cycle is adjusted back proportionally. Through nonlinear pre-prediction of individual pixel light decay and hybrid dynamic compensation of pulse width modulation and reference current, the brightness unevenness phenomenon caused by pixel-level light decay nonlinear differences during long-term operation of the display panel is eliminated.

[0046] 2. By using the real-time junction temperature as a thermal acceleration factor and constructing an exponential cross-product term with the cumulative driving pulse count and pulse width distribution parameters as cumulative degradation factors for normalization, the consistency between the expected brightness attenuation coefficient and the actual physical light decay process is improved. In the low attenuation range, the pulse width modulation duty cycle is adjusted separately to avoid display abnormalities caused by reference current fluctuations in the low grayscale range. In the high attenuation range, the reference value of the constant current source reference current is increased and the pulse width modulation duty cycle is adjusted back proportionally. This breaks through the compensation upper limit bottleneck caused by the scanning clock resolution limitation of single pulse width modulation adjustment in the high attenuation range, and ensures the consistency of attenuation compensation in the high and low grayscale ranges. Attached Figure Description

[0047] Figure 1 This is a flowchart illustrating the overall execution process of LED display screen brightness monitoring according to the present invention.

[0048] Figure 2 This is a flowchart of the real-time junction temperature acquisition and parasitic effect correction of the junction temperature sensing network of the present invention;

[0049] Figure 3 This is a flowchart of the driving historical statistical array parameter recording and weighting coefficient dynamic update process of the present invention;

[0050] Figure 4 This is a flowchart of the operation of the nonlinear coupling prediction model of optical decay-junction temperature-drive history of the present invention;

[0051] Figure 5 This is a flowchart of the low-attenuation range PWM duty cycle brightness compensation process of the present invention;

[0052] Figure 6 This is a flowchart of the high attenuation range reference current graded boost and PWM callback of the present invention. Detailed Implementation

[0053] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0054] Please refer to Figure 1 and Figure 2 This embodiment provides a method and system for monitoring the brightness of an LED display screen. The method acquires the real-time junction temperature of each LED pixel in the LED display screen through a junction temperature sensing network. The junction temperature sensing network consists of sensing circuits distributed within each pixel unit of the LED display screen, with each sensing circuit electrically connected to the corresponding LED pixel. During the row scanning cycle of the LED display screen, there is a fixed-length row blanking period, during which all LED pixels in that row are in the off state, and no forward drive current flows. During this row blanking period, the junction temperature sensing network initiates the junction temperature detection process, applying a reverse bias test pulse of a specific frequency to the row line where the LED pixel is located. The amplitude of the reverse bias test pulse is set to be less than the reverse breakdown voltage of the LED pixel and greater than the charging threshold voltage of the line parasitic capacitance. The duration of the reverse bias test pulse is set to be less than the total duration of the row blanking period to ensure that the junction temperature detection process is completed before the arrival of the forward drive pulse, without affecting the normal display scanning timing.

[0055] A junction temperature sensing network collects the transient voltage and current changes of the LED array under a reverse bias test pulse. The transient voltage change is obtained through a voltage sampling circuit connected in parallel across the LED array, employing a differential amplification structure to suppress common-mode noise interference. The transient current change is obtained through a sampling resistor connected in series in the LED array. The resistance value of the sampling resistor is chosen to be much smaller than the reverse impedance of the LED pixel to minimize its impact on the test pulse waveform. The sampling frequency of the sampling circuit is set to at least 10 times the frequency of the reverse bias test pulse to ensure complete capture of the transient voltage and current changes.

[0056] Based on the collected transient voltage and current changes, the dynamic impedance parameters of the LED pixel are calculated. The dynamic impedance parameters are calculated using the differential form of Ohm's law, whereby the dynamic impedance equals the ratio of the transient voltage change to the transient current change. The calculated dynamic impedance parameters are interpolated against a pre-calibrated impedance-temperature correlation curve, and the corresponding temperature value is extracted as the real-time junction temperature of the LED pixel. The impedance-temperature correlation curve is obtained by calibrating LED pixels of the same batch under constant temperature conditions. During calibration, the reverse dynamic impedance values ​​of the LED pixels at different temperatures are recorded, and a continuous impedance-temperature correlation curve is generated through polynomial fitting.

[0057] refer to Figure 3 The system records the cumulative number of driving pulses and pulse width distribution parameters for each LED pixel using a historical statistical array. This array consists of statistical units equal to the number of LED pixels, with each unit corresponding to one LED pixel and storing its cumulative driving pulse count and pulse width distribution parameters. The statistical units employ a non-volatile memory structure to ensure data is not lost when the display is powered off. Within each frame scan cycle, the historical statistical array counts the effective pulses that activate each LED pixel. The effective pulse count refers to the number of positive driving pulses applied to the LED pixel within that frame scan cycle. This effective pulse count is added to the cumulative driving pulse count corresponding to the LED pixel, reflecting the total number of times the LED pixel has been activated.

[0058] Simultaneously, the historical statistical array identifies the grayscale range of the forward drive pulse width corresponding to each LED pixel in each frame. The grayscale range is divided according to the grayscale levels of the LED display. For example, for an 8-bit grayscale LED display, it can be divided into eight grayscale ranges: 0-31, 32-63, 64-95, 96-127, 128-159, 160-191, 192-223, and 224-255. A corresponding weighting coefficient is assigned to each grayscale range. The magnitude of the weighting coefficient is positively correlated with the value of the grayscale range; that is, the weighting coefficient for a higher grayscale range is greater than that for a lower grayscale range. The forward drive pulse width is multiplied by the corresponding weighting coefficient and then summed to generate a weighted pulse width cumulative value. The weighted pulse width cumulative value and the cumulative drive pulse array together form the pulse width distribution parameter of the LED pixel, reflecting the driving time distribution of the LED pixel at different grayscale levels.

[0059] refer to Figure 4The model inputs real-time junction temperature, cumulative driving pulse count, and pulse width distribution parameters into a nonlinear coupled prediction model of light decay-junction temperature-driving history, outputting the expected brightness decay coefficient for each LED pixel. This model is a mathematical model built upon the physical mechanism of LED light decay, considering both the thermal acceleration effect of junction temperature and the cumulative degradation effect of driving history. Real-time junction temperature is used as the thermal acceleration factor, and the cumulative driving pulse count and pulse width distribution parameters are used as the cumulative degradation factor. An exponential cross-product term is constructed between the thermal acceleration factor and the cumulative degradation factor in the model. The value of this exponential cross-product term is calculated and normalized to obtain the expected brightness decay coefficient for each LED pixel. The expected brightness decay coefficient ranges from 0 to 1, where 1 indicates no brightness decay of the LED pixel and 0 indicates complete brightness decay. The expected brightness decay coefficient reflects the proportion of brightness decay of the current pixel relative to its initial brightness.

[0060] The mapping relationship between grayscale and current is reconstructed based on the expected brightness decay coefficient. The grayscale-current mapping relationship refers to the correspondence between the input grayscale data of the LED display and the actual output current of the LED pixels. At the factory, the initial grayscale-current mapping relationship is pre-calibrated, ensuring a linear correspondence between the input grayscale data and the output brightness when the LED pixels have not experienced light decay. When light decay occurs in the LED pixels, the grayscale-current mapping relationship needs to be reconstructed to compensate for the brightness loss caused by light decay.

[0061] refer to Figure 5 When the expected brightness attenuation coefficient is lower than a first set threshold, the constant current source reference current is kept constant, and the brightness is compensated by increasing the pulse width modulation duty cycle. The first set threshold is preset according to the performance requirements of the LED display and the adjustment characteristics of the constant current source, for example, it can be set to 0.8. When it is determined that the expected brightness attenuation coefficient is lower than the first set threshold, the reference voltage of the constant current source reference current is locked as the initial calibration value. The required brightness compensation gain value is calculated based on the expected brightness attenuation coefficient, and the brightness compensation gain value is equal to 1 divided by the expected brightness attenuation coefficient. The target grayscale data is determined based on the product of the brightness compensation gain value and the original grayscale data. The target grayscale data is remapped to the corresponding pulse width modulation duty cycle so that the actual output pulse width modulation duty cycle is greater than the duty cycle corresponding to the original grayscale data, thus completing the brightness compensation.

[0062] When the expected brightness attenuation coefficient exceeds a first preset threshold, the constant current source reference current is increased and the pulse width modulation duty cycle is proportionally adjusted back. When it is determined that the expected brightness attenuation coefficient exceeds the first preset threshold, the increase step size of the constant current source reference current is determined based on the attenuation range in which the expected brightness attenuation coefficient falls. The attenuation range is divided according to the first preset threshold and the maximum allowable output current of the constant current source; for example, it can be divided into multiple ranges such as 0.8-0.7, 0.7-0.6, and 0.6-0.5, each corresponding to a fixed increase step size. The reference voltage of the constant current source reference current is increased according to this increase step size, thereby increasing the output current of the constant current source. Simultaneously, a duty cycle adjustment coefficient is determined based on the ratio of the increased constant current source reference current to the initial calibration value. The duty cycle adjustment coefficient is equal to the initial calibration value divided by the increased constant current source reference current. The pulse width modulation duty cycle corresponding to the original grayscale data is multiplied by this duty cycle adjustment coefficient to generate an updated pulse width modulation duty cycle. By increasing the constant current source reference current and proportionally adjusting the pulse width modulation duty cycle, the requirement for the pulse width modulation duty cycle is reduced while ensuring that the total output brightness remains unchanged. This breaks through the bottleneck of the compensation limit caused by the scanning clock resolution limitation in the high attenuation range of single pulse width modulation adjustment.

[0063] In this embodiment, the brightness compensation methods under different expected brightness attenuation coefficients are compared as shown in Table 1.

[0064] Table 1 Comparison of brightness compensation methods under different expected brightness attenuation coefficients.

[0065]

[0066] Table 1 shows that when the expected brightness attenuation coefficient is in the range of 1.0-0.8, brightness compensation is achieved using pure pulse width modulation (PWM) duty cycle adjustment. When the expected brightness attenuation coefficient is below 0.8, the constant current source reference current is increased, and the PWM duty cycle is adjusted back proportionally. As the expected brightness attenuation coefficient decreases, the increase in the constant current source reference current gradually increases, and the adjustment of the PWM duty cycle also gradually increases. This hybrid adjustment method avoids display abnormalities caused by reference current fluctuations in the low grayscale range in the low attenuation range, and breaks through the compensation upper limit bottleneck of single PWM adjustment in the high attenuation range.

[0067] In this embodiment, the LED display brightness monitoring method acquires the junction temperature and driving history information of each LED pixel in real time, constructs a nonlinear coupling prediction model of light decay-junction temperature-driving history, pre-predicts the light decay of each LED pixel, and performs differentiated brightness compensation based on the prediction results using a hybrid adjustment method of pulse width modulation and reference current. This method can accurately compensate for the nonlinear light decay differences of individual pixels, eliminating the brightness unevenness caused by pixel-level nonlinear light decay differences during long-term operation of the display panel.

[0068] In a preferred embodiment, the process of acquiring the real-time junction temperature of each LED pixel in the LED display screen via a junction temperature sensing network also includes correction for parasitic effects on the circuitry. Before applying a reverse bias test pulse of a specific frequency, the junction temperature sensing network collects the initial parasitic capacitance voltage and leakage current of the horizontal lines. The initial parasitic capacitance voltage refers to the voltage existing on the horizontal lines due to charge accumulation when no reverse bias test pulse is applied. The leakage current refers to the small current existing between the horizontal lines and ground when no reverse bias test pulse is applied. The initial parasitic capacitance voltage and leakage current are collected before the start of each junction temperature detection process to ensure that the collected parameters are the parasitic parameters of the circuitry at the current moment.

[0069] Based on the acquired initial parasitic capacitance voltage and leakage current, the line parasitic effect offset is calculated. The line parasitic effect offset includes two parts: voltage offset and current offset. The voltage offset equals the initial parasitic capacitance voltage, and the current offset equals the leakage current. When calculating the dynamic impedance parameters, the voltage offset is subtracted from the acquired transient voltage change values ​​to generate corrected transient voltage change values; similarly, the current offset is subtracted from the acquired transient current change values ​​to generate corrected transient current change values. The corrected transient voltage change values ​​and corrected transient current change values ​​are used to calculate the dynamic impedance parameters, thereby eliminating the influence of line parasitic effects on junction temperature monitoring results.

[0070] Specifically, the calculation process for the line parasitic effect offset is as follows:

[0071] Let the initial parasitic capacitance voltage be Leakage current is The collected transient voltage change value is The collected transient current change value is The corrected transient voltage change value for:

[0072]

[0073] Corrected transient current change value for:

[0074]

[0075] Dynamic impedance parameters for:

[0076]

[0077] in, This is the corrected transient voltage change. This is the corrected transient current change.

[0078] The calculated dynamic impedance parameters Interpolation matching is performed with a pre-calibrated impedance-temperature correlation curve to extract the corresponding temperature value as the real-time junction temperature of the LED pixel. The impedance-temperature correlation curve is generated using cubic polynomial fitting, and its expression is:

[0079]

[0080] in, For the real-time junction temperature of LED pixels, , , , The coefficients are polynomial fitting coefficients, determined through calibration experiments.

[0081] In this embodiment, the reverse dynamic impedance calibration values ​​of the LED pixels at different temperatures are shown in Table 2.

[0082] Table 2. Reverse dynamic impedance calibration values ​​of LED pixels at different temperatures.

[0083]

[0084] The data in Table 2 were obtained by calibrating the same batch of LED pixels under constant temperature conditions. The real-time junction temperature of the LED pixel can be obtained by interpolating the calculated dynamic impedance parameters with the data in Table 2. For example, when the calculated dynamic impedance parameter is 90.0 kΩ, the corresponding temperature can be obtained as 42.3℃ through linear interpolation.

[0085] In this embodiment, by introducing line parasitic effect correction during junction temperature detection, the influence of line parasitic capacitance and leakage current on the calculation of dynamic impedance parameters is eliminated, thus improving the accuracy of junction temperature detection. Accurate junction temperature detection results are the basis for the optical decay prediction model to output accurate predicted brightness attenuation coefficients, thereby ensuring the accuracy of brightness compensation.

[0086] In a preferred embodiment, the process of recording the cumulative number of driving pulses and pulse width distribution parameters of each LED pixel through the driving historical statistics array further includes dynamic updating of the weighting coefficients. The cumulative operating time of the LED pixels is obtained. This cumulative operating time is recorded by a timer inside the driving historical statistics array. The timer starts counting when the LED display is powered on and stops counting when the power is off. The recording accuracy of the cumulative operating time is in minutes. Based on the cumulative operating time, a preset weighting coefficient attenuation table is consulted to obtain the attenuation correction factor corresponding to the cumulative operating time. The weighting coefficient attenuation table is pre-stored in the non-volatile memory of the driving historical statistics array. The attenuation correction factor ranges from 0 to 1 and decreases monotonically with the increase of the cumulative operating time.

[0087] The initially set weighting coefficients are multiplied by the attenuation correction factor to generate dynamically updated weighting coefficients. These dynamically updated weighting coefficients are then used in the calculation of the cumulative weighted pulse width value for the next frame. As the cumulative working time increases, the attenuation correction factor gradually decreases, causing the weight of the weighting coefficients corresponding to high grayscale ranges in the total pulse width distribution parameters to decrease over time. This is because the light decay rate of LED pixels at high grayscale levels gradually flattens out with increasing working time. By dynamically reducing the weighting coefficients in the high grayscale range, the pulse width distribution parameters can more accurately reflect the actual light decay of the LED pixels.

[0088] Specifically, the dynamic update process of the weighting coefficients is as follows:

[0089] Let the cumulative working time of the LED pixel be... The initial setting of the first The weighting coefficients for each grayscale interval are: The weighted coefficient decay table is related to the cumulative working time. The corresponding attenuation correction factor is Then the dynamically updated first... Weighting coefficients for each grayscale interval for:

[0090]

[0091] in, satisfy , and when hour, .

[0092] Within each frame scan cycle, let the first... The forward driving pulse width corresponding to each grayscale interval is The weighted pulse width increment of that frame for:

[0093]

[0094] in, This represents the total number of grayscale intervals.

[0095] The weighted pulse width increment Accumulated to weighted pulse width cumulative value In, that is:

[0096]

[0097] Cumulative number of driving pulses The update process is as follows:

[0098]

[0099] in, This is the count of effective pulses that turn on the LED pixel within the scanning period of that frame.

[0100] Pulse width distribution parameters are derived from weighted cumulative pulse width values. and cumulative number of driving pulses Combining, that is:

[0101]

[0102] In this embodiment, the weighted coefficient attenuation correction factor for different cumulative working times is shown in Table 3.

[0103] Table 3 Weighting coefficient attenuation correction factor under different cumulative working times

[0104]

[0105] The data in Table 3 were obtained through long-term aging experiments on the same batch of LED pixels. The experimental results show that as the cumulative working time increases, the light decay rate of the LED pixels at high grayscale levels gradually decreases. Therefore, it is necessary to correspondingly reduce the weighting coefficients in the high grayscale range. By dynamically updating the weighting coefficients, the pulse width distribution parameters can more accurately reflect the actual light decay of the LED pixels, thereby improving the accuracy of the light decay prediction model.

[0106] In this embodiment, by introducing a dynamic update mechanism for weighted coefficients, the characteristics of LED pixel light decay rate changing with operating time are considered, enabling the pulse width distribution parameters to more accurately reflect the impact of the actual driving history of the LED pixel on light decay. Accurate pulse width distribution parameters are crucial for the light decay prediction model to output accurate predicted brightness attenuation coefficients, thereby further improving the accuracy of brightness compensation.

[0107] In a preferred embodiment, the process of inputting real-time junction temperature, cumulative driving pulse count, and pulse width distribution parameters into the nonlinear coupling prediction model of light decay-junction temperature-driving history, and outputting the expected brightness decay coefficient of each LED pixel, includes constructing an exponential cross-product term of the thermal acceleration factor and the cumulative degradation factor. The nonlinear coupling prediction model of light decay-junction temperature-driving history is based on the Arrhenius equation and the cumulative damage theory. This model considers the thermal acceleration effect of junction temperature on light decay and the cumulative degradation effect of driving history on light decay, while also considering the nonlinear coupling relationship between the thermal acceleration effect and the cumulative degradation effect.

[0108] Real-time junction temperature As a thermal accelerator The expression for the thermal acceleration factor is based on the Arrhenius equation:

[0109]

[0110] in, The activation energy for LED light decay, Boltzmann's constant, This represents the real-time junction temperature of the LED pixel, measured in Kelvin.

[0111] Accumulated drive pulse count Weighted cumulative pulse width in pulse width distribution parameters As a cumulative degradation factor The expression for the cumulative degradation factor is:

[0112]

[0113] in, and These are model coefficients, determined through aging experiments. This reflects the impact of the cumulative number of driving pulses on optical decay. This reflects the impact of the weighted pulse width accumulation on light decay.

[0114] In the nonlinear coupling prediction model of light decay-junction temperature-drive history, an exponential cross-product term of the thermal acceleration factor and the cumulative degradation factor is constructed. This cross-product term reflects the nonlinear coupling relationship between the thermal acceleration effect and the cumulative degradation effect. (Expected brightness decay coefficient) The expression is:

[0115]

[0116] in, The coupling coefficient is determined through aging experiments.

[0117] The calculated expected brightness attenuation coefficient is normalized to a value between 0 and 1. The normalization process is as follows:

[0118]

[0119] in, This is the expected brightness decay coefficient when the LED pixel does not experience light decay, and its value is 1. This is the expected brightness decay coefficient when the LED pixel fully decays, and its value is 0. Therefore, the normalized expected brightness decay coefficient is... Equal to the original expected brightness attenuation coefficient .

[0120] In this embodiment, the coefficients of the nonlinear coupling prediction model of light decay-junction temperature-drive history are determined by conducting accelerated aging experiments on the same batch of LED pixels. The accelerated aging experiments are conducted under different junction temperature conditions and different drive conditions, recording the brightness decay of the LED pixels, and the coefficients of the model are obtained by fitting the model using the least squares method. , , and During the fitting process, the objective function is to minimize the mean square error between the actual measured brightness attenuation coefficient and the expected brightness attenuation coefficient predicted by the model.

[0121] In this embodiment, by constructing an exponential cross-product term of the thermal acceleration factor and the cumulative degradation factor, the nonlinear coupling relationship between junction temperature and driving history is considered, enabling the light decay prediction model to more accurately reflect the actual light decay process of LED pixels. An accurate expected brightness decay coefficient is a prerequisite for precise brightness compensation, thereby ensuring the brightness uniformity of the display panel after long-term operation.

[0122] In a preferred embodiment, when the expected brightness attenuation coefficient exceeds a first preset threshold, the process of increasing the constant current source reference current reference value and proportionally adjusting the pulse width modulation duty cycle further includes a graded, progressive increase of the reference current. After increasing the reference voltage of the constant current source reference current in the row containing the target LED pixel, the reference voltage difference between the target LED pixel and adjacent LED pixels is calculated. Adjacent LED pixels include pixels adjacent to the target LED pixel in the top, bottom, left, and right directions. If the reference voltage difference exceeds a preset voltage step threshold, the increase step size of the target LED pixel is graded and split, replacing the single increase step size with a multi-level progressive increase step size. The voltage step threshold is preset based on the sensitivity of the human eye to brightness changes, for example, it can be set to 5% of the initial reference voltage.

[0123] Over multiple consecutive frame cycles, the constant current source reference current value of the target LED pixel is progressively increased in stages with multi-level incremental increments, while the duty cycle callback coefficient is adjusted synchronously at each stage. Each increment has the same step size, and each increment is completed within one frame cycle. By progressively increasing the constant current source reference current in stages, brightness jumps caused by sudden changes in the reference current can be avoided, thereby improving the visual comfort of the display.

[0124] Specifically, the stepwise increase process of the reference current is as follows:

[0125] Let the expected brightness attenuation coefficient of the target LED pixel be... The first set threshold is .when At that time, the target boost ratio of the constant current source reference current is determined to be: The initial reference current is The target reference current is .

[0126] Calculate the maximum reference voltage difference between the target LED pixel and its adjacent LED pixels. .like ,in If the voltage step threshold is used, then the boost step size is split into... Level, the percentage increase for each level is . The value of satisfies The smallest integer of , where This is the switching resistor between the constant current source reference voltage and the reference current.

[0127] In the Within each frame period, the boosted reference current is:

[0128]

[0129] in, .

[0130] The corresponding duty cycle callback coefficient is:

[0131]

[0132] The pulse width modulation duty cycle corresponding to the original grayscale data Multiply by the duty cycle callback factor Generate the updated pulse width modulation duty cycle :

[0133]

[0134] In this embodiment, by introducing a graded, progressively increasing reference current mechanism, brightness jumps caused by excessive differences in reference current between adjacent pixels are avoided, thus improving visual comfort. Simultaneously, the duty cycle callback coefficient is adjusted synchronously at each level to ensure that the output brightness of the LED pixels remains stable during the reference current increase process, preventing brightness fluctuations.

[0135] In a preferred embodiment, the LED display brightness monitoring system includes a junction temperature sensing network, a drive history statistics array, a light decay prediction arithmetic unit, and a hybrid drive logic controller. The junction temperature sensing network is electrically connected to each LED pixel of the LED display and is used to acquire the real-time junction temperature of each LED pixel. The junction temperature sensing network includes multiple row sensing units, each corresponding to one row of pixels on the LED display. These units are used to apply a reverse bias test pulse during the row blanking period and to collect the transient voltage and current changes of the row line. Each row sensing unit includes a pulse generator, a voltage sampling circuit, a current sampling circuit, and an impedance calculation circuit. The pulse generator generates a reverse bias test pulse of a specific frequency; the voltage and current sampling circuits collect the transient voltage and current changes of the row line, respectively; and the impedance calculation circuit calculates the dynamic impedance parameters of the LED pixel based on the collected transient voltage and current changes and converts the dynamic impedance parameters into a real-time junction temperature.

[0136] The driving history statistics array is connected to the driving circuit of the LED display screen and is used to record the cumulative driving pulse count and pulse width distribution parameters of each LED pixel. The driving history statistics array includes multiple pixel statistics units, each corresponding to one LED pixel, used to store the cumulative driving pulse count, weighted cumulative pulse width value, and cumulative working time of that pixel. Each pixel statistics unit includes a pulse counter, a pulse width integrator, a weighting arithmetic unit, and a timer. The pulse counter counts the effective pulses that turn on the LED pixel within each frame scan cycle and adds them to the cumulative driving pulse count; the pulse width integrator measures the forward driving pulse width corresponding to the LED pixel in each frame; the weighting arithmetic unit multiplies the forward driving pulse width by the corresponding weighting coefficient and adds it to the weighted cumulative pulse width value, dynamically updating the weighting coefficients according to the cumulative working time; the timer records the cumulative working time of the LED pixel.

[0137] The light decay prediction calculator is connected to both the junction temperature sensing network and the drive history statistical array. It inputs real-time junction temperature, cumulative drive pulse count, and pulse width distribution parameters into the light decay-junction temperature-drive history nonlinear coupled prediction model, outputting the expected brightness decay coefficient for each LED pixel. The light decay prediction calculator includes a data preprocessing unit, a model calculation unit, and a result output unit. The data preprocessing unit performs format conversion and normalization on the real-time junction temperature data output from the junction temperature sensing network and the cumulative drive pulse count and pulse width distribution parameters output from the drive history statistical array. The model calculation unit incorporates the light decay-junction temperature-drive history nonlinear coupled prediction model, used to calculate the expected brightness decay coefficient for each LED pixel based on the preprocessed data. The result output unit outputs the calculated expected brightness decay coefficient to the hybrid drive logic controller.

[0138] The hybrid drive logic controller is connected to both the light decay prediction arithmetic unit and the constant current source drive circuit of the LED display screen, and is used to reconstruct the mapping relationship between grayscale and current based on the expected brightness decay coefficient. The hybrid drive logic controller includes a threshold judgment unit, a constant current source control unit, and a pulse width modulation control unit. The threshold judgment unit compares the expected brightness decay coefficient with a first set threshold to determine the decay range of the current LED pixel; the constant current source control unit controls the reference voltage of the constant current source reference current based on the output of the threshold judgment unit; and the pulse width modulation control unit adjusts the pulse width modulation duty cycle based on the output of the threshold judgment unit and the change in the constant current source reference current.

[0139] When the expected brightness attenuation coefficient is lower than a first set threshold, the threshold judgment unit outputs a first control signal to the constant current source control unit and the pulse width modulation control unit. Upon receiving the first control signal, the constant current source control unit locks the reference voltage of the constant current source reference current as the initial calibration value. Upon receiving the first control signal, the pulse width modulation control unit calculates the brightness compensation gain value based on the expected brightness attenuation coefficient, determines the target grayscale data based on the product of the brightness compensation gain value and the original grayscale data, and remaps the target grayscale data to the corresponding pulse width modulation duty cycle.

[0140] When the expected brightness attenuation coefficient exceeds a first preset threshold, the threshold judgment unit outputs a second control signal to the constant current source control unit and the pulse width modulation control unit. Upon receiving the second control signal, the constant current source control unit determines the boost step size of the constant current source reference current based on the attenuation range where the expected brightness attenuation coefficient lies, and increases the reference voltage of the constant current source reference current according to this boost step size. Simultaneously, upon receiving the second control signal, the pulse width modulation control unit determines the duty cycle callback coefficient based on the ratio of the boosted constant current source reference current to the initial calibration value, multiplies the pulse width modulation duty cycle corresponding to the original grayscale data by this duty cycle callback coefficient, and generates an updated pulse width modulation duty cycle.

[0141] In this embodiment, the various modules of the LED display brightness monitoring system work collaboratively to achieve real-time monitoring, pre-prediction, and differentiated compensation of light decay in each pixel of the LED display. The junction temperature sensor network acquires the junction temperature information of each pixel in real time, drives the historical statistical array to record the driving history information of each pixel, and the light decay prediction calculator predicts the expected brightness decay coefficient of each pixel based on the junction temperature information and driving history information. The hybrid driving logic controller performs brightness compensation using a hybrid adjustment method of pulse width modulation and reference current based on the expected brightness decay coefficient. This system can accurately compensate for the nonlinear light decay differences of individual pixels, eliminating the brightness unevenness caused by pixel-level nonlinear differences in light decay during long-term operation of the display panel.

[0142] In a preferred embodiment, the row sensing unit of the junction temperature sensing network further includes a parasitic effect correction circuit. The parasitic effect correction circuit is connected to the voltage sampling circuit and the current sampling circuit, and is used to acquire the initial parasitic capacitance voltage and leakage current of the row line before applying the reverse bias test pulse, and calculate the line parasitic effect offset based on the initial parasitic capacitance voltage and leakage current. When calculating the dynamic impedance parameters, the parasitic effect correction circuit subtracts the voltage offset from the transient voltage change value output by the voltage sampling circuit and subtracts the current offset from the transient current change value output by the current sampling circuit, generating corrected transient voltage change values ​​and transient current change values. The impedance calculation circuit uses the corrected transient voltage change values ​​and corrected transient current change values ​​to calculate the dynamic impedance parameters, thereby eliminating the influence of line parasitic effects on the junction temperature detection results.

[0143] In a preferred embodiment, the pixel statistics unit driving the historical statistics array further includes a weighting coefficient update circuit. The weighting coefficient update circuit is connected to a timer and a weighting arithmetic unit. It is used to look up a preset weighting coefficient attenuation table based on the cumulative working time recorded by the timer, obtain the corresponding attenuation correction factor, and multiply the initially set weighting coefficient by the attenuation correction factor to generate dynamically updated weighting coefficients. The weighting arithmetic unit uses the dynamically updated weighting coefficients to calculate the weighted pulse width increment and accumulates the weighted pulse width increment into the cumulative weighted pulse width value. By dynamically updating the weighting coefficients, the pulse width distribution parameters can more accurately reflect the actual light decay of the LED pixels.

[0144] In a preferred embodiment, reference Figure 6 The hybrid drive logic controller also includes a graded boost control unit. This unit, connected to the constant current source control unit and the pulse width modulation control unit, calculates the reference voltage difference between the target LED pixel and its adjacent LED pixels after boosting the reference voltage of the constant current source reference current in the row containing the target LED pixel. If this reference voltage difference exceeds a set voltage step threshold, the graded boost control unit divides the boost step size of the target LED pixel into graded steps, replacing a single boost step size with a multi-level progressive boost step size. Over multiple consecutive frame cycles, the graded boost control unit controls the constant current source control unit to progressively boost the constant current source reference current value according to the multi-level progressive boost step size, while simultaneously controlling the pulse width modulation control unit to synchronously adjust the duty cycle callback coefficient. By progressively boosting the constant current source reference current, brightness jumps caused by sudden changes in the reference current are avoided, improving the visual comfort of the display.

[0145] In a preferred embodiment, the model computation unit of the light decay prediction calculator further includes a model parameter update module. This module is connected to an external aging test device and receives LED pixel brightness decay data output by the device. Based on this data, it updates the coefficients of the nonlinear coupling prediction model for light decay, junction temperature, and drive history. The module employs online learning to continuously optimize the model coefficients, enabling the model to more accurately reflect the actual light decay process of the LED pixels. By updating the model parameters online, the adaptability and accuracy of the light decay prediction model are improved.

[0146] In a preferred embodiment, the LED display brightness monitoring system further includes a data storage unit. The data storage unit is connected to a junction temperature sensing network, a drive history statistics array, a light decay prediction calculator, and a hybrid drive logic controller, respectively, and is used to store junction temperature detection data, drive history data, light decay prediction data, and brightness compensation data. The data storage unit employs a non-volatile memory structure to ensure that data is not lost after the display is powered off. By storing this data, the operating status of the LED display can be monitored and analyzed over a long period, providing data support for the maintenance and upkeep of the LED display.

[0147] In a preferred embodiment, the LED display brightness monitoring system further includes a communication interface unit. The communication interface unit is connected to the data storage unit and is used to upload the data stored in the data storage unit to a remote monitoring platform. The remote monitoring platform can centrally monitor and manage the operating status of multiple LED displays, promptly detect abnormalities, and issue alarm signals. Through the communication interface unit and the remote monitoring platform, remote monitoring and intelligent management of the LED displays are achieved.

Claims

1. A method for monitoring the brightness of an LED display screen, characterized in that, include: The real-time junction temperature of each LED pixel in the LED display screen is obtained through a junction temperature sensing network. The cumulative number of driving pulses and pulse width distribution parameters of each LED pixel are recorded by driving a historical statistical array. The real-time junction temperature, the cumulative number of driving pulses, and the pulse width distribution parameters are input into the nonlinear coupling prediction model of light decay-junction temperature-driving history, and the expected brightness decay coefficient of each LED pixel is output. Based on the expected brightness attenuation coefficient, the mapping relationship between grayscale and current is reconstructed. When the expected brightness attenuation coefficient is lower than the first set threshold, the constant current source reference current is kept unchanged, and the brightness is compensated by increasing the pulse width modulation duty cycle. When the expected brightness attenuation coefficient exceeds the first set threshold, the pulse width modulation duty cycle is adjusted back proportionally.

2. The LED display screen brightness monitoring method according to claim 1, characterized in that, The method of obtaining the real-time junction temperature of each LED pixel in the LED display screen through the junction temperature sensing network includes: during the line scan blanking period, applying a reverse bias test pulse of a specific frequency to the line line where the LED pixel is located, and keeping the duration of the reverse bias test pulse less than the total duration of the line blanking period. The transient voltage and transient current changes of the line under the reverse bias test pulse are collected. The dynamic impedance parameters of the LED pixel are calculated based on the transient voltage change value and the transient current change value; The dynamic impedance parameter is interpolated and matched with the pre-calibrated impedance-temperature correlation curve, and the temperature value corresponding to the dynamic impedance parameter is extracted as the real-time junction temperature of each LED pixel.

3. The LED display screen brightness monitoring method according to claim 1, characterized in that, The method of recording the cumulative number of driving pulses and pulse width distribution parameters of each LED pixel through the driving history statistical array includes: for each LED pixel, counting the effective pulse count of the LED pixel being turned on in each frame scan cycle; The effective pulse count is added to the cumulative drive pulse count; Simultaneously, the grayscale range of the positive driving pulse width corresponding to the LED pixel in each frame is identified, and corresponding weighting coefficients are set for different grayscale ranges; The positive driving pulse width is multiplied by the corresponding weighting coefficient and then accumulated to generate a weighted pulse width cumulative value. The weighted pulse width cumulative value and the cumulative driving pulse array are combined to form the pulse width distribution parameter.

4. The LED display screen brightness monitoring method according to claim 1, characterized in that, The real-time junction temperature, the cumulative number of driving pulses, and the pulse width distribution parameters are input into the nonlinear coupling prediction model of light decay-junction temperature-driving history, and the expected brightness decay coefficient of each LED pixel is output, including: using the real-time junction temperature as a thermal acceleration factor, and using the cumulative number of driving pulses and the pulse width distribution parameters as a cumulative degradation factor. In the nonlinear coupling prediction model of optical decay-junction temperature-drive history, an exponential cross-product term is constructed between the thermal acceleration factor and the cumulative degradation factor. The value of the exponential cross product term is calculated and normalized to obtain the expected brightness decay coefficient of each LED pixel. The expected brightness decay coefficient reflects the decay ratio of the current pixel brightness relative to the initial brightness.

5. The LED display screen brightness monitoring method according to claim 1, characterized in that, When the expected brightness attenuation coefficient is lower than the first set threshold, the constant current source reference current is kept unchanged, and the brightness is compensated by increasing the pulse width modulation duty cycle, including: when it is determined that the expected brightness attenuation coefficient is lower than the first set threshold, the reference voltage of the constant current source reference current is locked as the initial calibration value; Calculate the required brightness compensation gain value based on the expected brightness attenuation coefficient; The target grayscale data is determined based on the product of the brightness compensation gain value and the original grayscale data. The target grayscale data is remapped to the corresponding pulse width modulation duty cycle so that the actual output pulse width modulation duty cycle is greater than the duty cycle corresponding to the original grayscale data, thus completing the brightness compensation.

6. The LED display screen brightness monitoring method according to claim 1, characterized in that, When the expected brightness attenuation coefficient exceeds the first set threshold, the pulse width modulation duty cycle is proportionally reverted, which includes: when it is determined that the expected brightness attenuation coefficient exceeds the first set threshold, determining the boosting step size of the constant current source reference current according to the attenuation range in which the expected brightness attenuation coefficient is located, and increasing the reference voltage of the constant current source reference current according to the boosting step size; Meanwhile, the duty cycle callback coefficient is determined based on the ratio of the enhanced constant current source reference current to the initial calibration value. The pulse width modulation duty cycle corresponding to the original grayscale data is multiplied by the duty cycle callback coefficient to generate the updated pulse width modulation duty cycle.

7. The LED display screen brightness monitoring method according to claim 2, characterized in that, The acquisition of transient voltage and transient current changes of the line under the reverse bias test pulse also includes: acquiring the initial parasitic capacitance voltage and leakage current of the line before applying the reverse bias test pulse of the specific frequency; The line parasitic effect offset is calculated based on the initial parasitic capacitance voltage and leakage current. When calculating the dynamic impedance parameters, the line parasitic effect offset is subtracted from the transient voltage change value and transient current change value to generate corrected transient voltage change value and transient current change value, and the dynamic impedance parameters are calculated using the corrected values.

8. The LED display screen brightness monitoring method according to claim 3, characterized in that, The step of setting corresponding weighting coefficients for different grayscale ranges also includes: obtaining the cumulative working time of the LED pixel during continuous operation; Based on the cumulative working time, a preset weighted coefficient attenuation table is searched to obtain the attenuation correction factor corresponding to the cumulative working time; The weighting coefficients initially set are multiplied by the attenuation correction factor to generate dynamically updated weighting coefficients; As the cumulative working time increases, the attenuation correction factor decreases monotonically, causing the weight of the weighting coefficient corresponding to the high grayscale interval in the total pulse width distribution parameter to decrease over time. The dynamically updated weighting coefficient participates in the calculation of the weighted pulse width cumulative value in the next frame.

9. The LED display screen brightness monitoring method according to claim 6, characterized in that, The reference voltage for increasing the constant current source reference current according to the aforementioned step size also includes: after increasing the reference voltage of the constant current source reference current in the row where the target LED pixel is located, calculating the reference voltage difference between the target LED pixel and the adjacent LED pixel. If the reference voltage difference exceeds the set voltage step threshold, the lifting step size of the target LED pixel is divided into levels, and the single lifting step size is replaced with a multi-level progressive lifting step size. The duty cycle callback coefficient is adjusted synchronously and progressively over multiple consecutive frame periods.

10. An LED display screen brightness monitoring system, characterized in that, include: Junction temperature sensing network is used to obtain the real-time junction temperature of each LED pixel in the LED display screen; A historical statistics array is used to record the cumulative number of driving pulses and pulse width distribution parameters of each LED pixel; The light decay prediction calculator is used to input the real-time junction temperature, the cumulative number of driving pulses and the pulse width distribution parameters into the light decay-junction temperature-driving history nonlinear coupling prediction model, and output the expected brightness decay coefficient of each LED pixel. A hybrid drive logic controller is used to reconstruct the mapping relationship between grayscale and current based on the expected brightness attenuation coefficient. When the expected brightness attenuation coefficient is lower than a first set threshold, the constant current source reference current is kept unchanged, and the brightness is compensated by increasing the pulse width modulation duty cycle. When the expected brightness attenuation coefficient exceeds the first set threshold, the pulse width modulation duty cycle is adjusted back proportionally.

Images