Method and device for optimizing driving waveform of small-pitch LED array

Abstract

The present invention relates to the technical field of LED array driving, and discloses a method and device for optimizing the driving waveform of a small-pitch LED array. The method includes: collecting the spectral characteristics of pixel points in the small-pitch LED array to construct a mathematical model of the luminous characteristics of pixel points; performing parameter operations on the mathematical model of the luminous characteristics of pixel points to obtain array display parameters, and constructing a display quality evaluation function according to the array display parameters; performing waveform compensation on the original PWM driving signal to generate a pair of complementary driving waveforms, and performing an orthogonal test based on the pair of complementary driving waveforms to establish a characteristic correlation matrix between driving parameters and display effects; calculating an optimal combination of driving parameters according to the characteristic correlation matrix and the display quality evaluation function, and generating an LED array driving signal according to the optimal combination of driving parameters. The present invention realizes high-precision control of the driving waveform and ensures the stability and reliability of the driving signal.

Description

Technical Field

[0001] The present invention relates to the technical field of LED array driving, and particularly to a method and device for optimizing the driving waveform of a small-pitch LED array. Background Art

[0002] During the driving process of a small-pitch LED array, due to the continuous reduction of the pixel pitch, the crosstalk effect between adjacent pixels becomes increasingly significant, which not only reduces the clarity of the display image but also affects the overall performance of the display system.

[0003] The traditional PWM driving method has many problems when dealing with small-pitch LED arrays: overshoot and ringing are likely to occur at the rising and falling edges of the driving waveform, resulting in waveform distortion; secondly, the simultaneous conduction of adjacent pixels will cause power supply ripple, resulting in crosstalk effect; thirdly, due to the lack of an effective driving waveform optimization mechanism, it is difficult to achieve high gray-scale control accuracy while ensuring display uniformity. Summary of the Invention

[0004] The present invention provides a method and device for optimizing the driving waveform of a small-pitch LED array, which realizes high-precision control of the driving waveform and ensures the stability and reliability of the driving signal.

[0005] In the first aspect, the present invention provides a method for optimizing the driving waveform of a small-pitch LED array, and the method for optimizing the driving waveform of a small-pitch LED array includes:

[0006] Collect the spectral characteristics of the pixels in the small-pitch LED array and construct a mathematical model of the light-emitting characteristics of the pixels;

[0007] Perform parameter operations on the mathematical model of the light-emitting characteristics of the pixels to obtain array display parameters, and construct a display quality evaluation function according to the array display parameters;

[0008] Perform waveform compensation on the original PWM driving signal to generate a pair of complementary driving waveforms, and perform an orthogonal test based on the pair of complementary driving waveforms to establish a characteristic correlation matrix between the driving parameters and the display effect;

[0009] Calculate the optimal driving parameter combination according to the characteristic correlation matrix and the display quality evaluation function, and generate an LED array driving signal according to the optimal driving parameter combination.

[0010] In the second aspect, the present invention provides a device for optimizing the driving waveform of a small-pitch LED array, and the device for optimizing the driving waveform of a small-pitch LED array includes:

[0011] An acquisition module for collecting the spectral characteristics of the pixels in the small-pitch LED array and constructing a mathematical model of the light-emitting characteristics of the pixels;

[0012] A building block for performing parameter operations on the mathematical model of the light-emitting characteristics of the pixel points to obtain array display parameters, and constructing a display quality evaluation function according to the array display parameters;

[0013] A testing module for performing waveform compensation on the original PWM driving signal to generate a complementary driving waveform pair, and performing an orthogonal test based on the complementary driving waveform pair to establish a characteristic correlation matrix between the driving parameters and the display effect;

[0014] A generating module for calculating an optimal driving parameter combination according to the characteristic correlation matrix and the display quality evaluation function, and generating an LED array driving signal according to the optimal driving parameter combination.

[0015] In the technical solution provided by the present invention, by establishing a complete mathematical model of the light-emitting characteristics of pixel points and combining temperature characteristic compensation, an accurate description of the display characteristics of the LED array is achieved. The design scheme of the complementary driving waveform pair is adopted, and by injecting positive and negative compensation pulses into the row and column scanning signals, the voltage spikes and crosstalk effects during the driving process are effectively suppressed, and the uniformity of the display screen is improved. An introduction of a method for establishing a characteristic correlation matrix based on orthogonal testing, through systematic parameter testing and data analysis, an accurate mapping relationship between the driving parameters and the display effect is established. An improved particle swarm optimization algorithm is designed, and through chaotic sequence initialization and adaptive weight adjustment, the convergence efficiency and solution quality of the optimization process are improved. A driving signal generation scheme combining a digital phase-locked loop and digital pulse width modulation is adopted to achieve high-precision control of the driving waveform and ensure the stability and reliability of the driving signal. Description of the Drawings

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0017] Figure 1 It is a schematic flow chart of a method for optimizing the driving waveform of a small-pitch LED array provided by an embodiment of the present application;

[0018] Figure 2 It is a schematic structural block diagram of a device for optimizing the driving waveform of a small-pitch LED array provided by an embodiment of the present application. Detailed Embodiments

[0019] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0020] The flowchart shown in the accompanying drawings is only an example for illustration, and does not necessarily include all the content and operations / steps, nor does it necessarily need to be executed in the described order. For example, some operations / steps can also be decomposed, combined, or partially merged, so the actual execution order may change based on the actual situation.

[0021] It should also be understood that the terms used in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application. As used in the specification of this application and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an", and "the" are intended to include the plural forms.

[0022] It should be further understood that the term "and / or" used in the specification of this application and the appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0023] Next, in conjunction with the accompanying drawings, some embodiments of this application will be described in detail. Without conflict, the embodiments and features in the following embodiments can be combined with each other.

[0024] Please refer to Figure 1 , Figure 1 which is a schematic flowchart of the method for optimizing the driving waveform of a small-pitch LED array provided by an embodiment of this application. As Figure 1 shown, the method for optimizing the driving waveform of a small-pitch LED array provided by an embodiment of this application includes steps S100 to S600.

[0025] Step S100: Collect the spectral characteristics of the pixel points in the small-pitch LED array and construct a mathematical model of the light-emitting characteristics of the pixel points;

[0026] It can be understood that the execution subject of the present invention can be a device for optimizing the driving waveform of a small-pitch LED array, or a terminal or a server. Specifically, it is not limited here. In the embodiments of the present invention, the server is used as the execution subject for illustration.

[0027] Specifically, the ambient temperature of the small-pitch LED array is detected. The luminous characteristics of the LED are significantly affected by temperature. Temperature changes will cause spectral wavelength drift and brightness fluctuations. The working ambient temperature of the LED array is monitored in real time through a temperature sensor or an infrared temperature measurement device, and corresponding temperature acquisition data is obtained. Based on the distribution characteristics of temperature, the current sampling interval is divided, so that the subsequent current input can more precisely match the working characteristics of the LED at different temperatures, in order to improve the accuracy of the model. Different driving currents are input to each pixel point in the LED array in sequence according to the set interval range, so as to observe the spectral response of the pixel point under different current conditions. A high-precision spectral analyzer is used to measure the luminous characteristics of each pixel point and record the spectral data, including spectral wavelength data and spectral intensity data. The spectral wavelength data reflects the main emission wavelength of the LED, while the spectral intensity data characterizes the luminous brightness of the pixel point under a specific current. Since the spectral characteristics of the LED change under different driving currents, measurements are carried out within the full current range to ensure the integrity and accuracy of the spectral data. The main wavelength extraction operation is performed on the spectral wavelength data to determine the main emission wavelength of the LED under different driving currents. The main wavelength is determined by calculating the centroid or peak position of the spectral curve, and the current-wavelength characteristic data is established at different currents, which describes how the main wavelength of the LED drifts with the change of the driving current. At the same time, for the spectral intensity data, integral operation is performed to calculate the total light output under different current conditions, that is, the current-brightness characteristic data, which can reflect the change of the luminous brightness of the LED with the driving current. When constructing the mathematical model of the luminous characteristics, mathematical expressions are established for the current-wavelength characteristic data and the current-brightness characteristic data respectively. Based on the current-wavelength characteristic data, a wavelength characteristic polynomial function is established. By selecting an appropriate polynomial order to fit the trend of the main wavelength changing with the current, and using the least squares fitting algorithm to calculate the coefficient matrix of the polynomial, the wavelength characteristic coefficient matrix is obtained. Similarly, for the current-brightness characteristic data, a brightness characteristic polynomial function is established, and the brightness characteristic coefficient matrix is solved through the least squares fitting operation. Based on the wavelength characteristic coefficient matrix and the brightness characteristic coefficient matrix, a three-dimensional characteristic surface is established, so as to mathematically describe the luminous characteristics of the LED pixel point under different current inputs. This three-dimensional characteristic surface takes the current as the independent variable and the wavelength and brightness as the response variables, and constructs a mathematical model of the pixel point luminous characteristics through the method of surface fitting, so that it can accurately reflect the luminous behavior of the LED under different driving conditions.

[0028] Step S200: Perform parameter operations on the mathematical model of the pixel point luminous characteristics to obtain the array display parameters, and construct a display quality evaluation function according to the array display parameters;

[0029] Specifically, parameter operations are performed on the mathematical model of the pixel light-emitting characteristics to obtain the key display parameters of the array. Among them, the average brightness data is the core index to measure the light-emitting uniformity of pixel points. Therefore, numerical integration operations need to be performed on the mathematical model to calculate the average brightness data of each pixel point, and the brightness uniformity of the entire array is calculated based on these data. The calculation method of the brightness uniformity coefficient is to determine the ratio of the maximum brightness value to the minimum brightness value, which can effectively reflect whether the brightness distribution of the LED array in different regions is uniform. Based on the analysis of brightness uniformity, Fourier transform is performed on the waveform response parameters in the mathematical model of pixel light-emitting characteristics to obtain the spectral component data of the driving signal. Fourier transform can convert the time-domain information of the PWM waveform into frequency-domain information, enabling the system to analyze whether there are high harmonic components in the driving signal and evaluating the waveform distortion rate by calculating the harmonic distortion factor. The level of waveform distortion rate directly affects the visual display effect of the LED. An excessive distortion rate will cause problems such as flicker, color deviation, or gray-scale instability. Therefore, it needs to be strictly controlled during the optimization process. Due to the optical crosstalk between pixel points in the small-pitch LED array, that is, the light radiation of adjacent pixel points will affect each other's chromaticity and brightness performance, an optical crosstalk model between pixel points is established based on the mathematical model of pixel light-emitting characteristics, and finite element analysis is performed on this model to simulate the light energy propagation path between different pixel points and calculate the influence range and attenuation characteristics of optical crosstalk. Through finite element analysis, the crosstalk attenuation ratio is obtained, which can quantitatively characterize the degree of optical interference between adjacent pixel points. A weight matrix is constructed based on the brightness uniformity coefficient, waveform distortion rate, and crosstalk attenuation ratio, and eigenvalue decomposition operations are used to solve the optimal weight coefficients. The construction process of the weight matrix is based on the actual requirements of the display system and the human eye visual sensitivity to ensure that the influence weights of different parameters are reasonably allocated in the comprehensive evaluation. The role of eigenvalue decomposition is to extract the main influencing factors, making the optimization process more stable and reducing the influence of redundant parameters. Normalization processing is performed on the brightness uniformity coefficient, waveform distortion rate, and crosstalk attenuation ratio to obtain standardized evaluation indicators. The standardized evaluation indicators are linearly weighted using the weight coefficients to obtain the initial evaluation scores. The hyperbolic tangent function is used to perform transformation operations on the initial evaluation scores to enhance the stability of the evaluation scores and ensure their variation within a certain range. The hyperbolic tangent function has good non-linear mapping characteristics, mapping the input value to a bounded interval, making the final display quality evaluation function more in line with the subjective perception of the human eye. This evaluation function is used as the objective function for driving waveform optimization, thereby guiding the adjustment of the PWM driving signal to make the LED array reach the optimal state in terms of brightness uniformity, waveform stability, and optical crosstalk suppression, thus significantly improving the overall visual effect of the LED display screen.

[0030] Step S300: Perform waveform compensation on the original PWM drive signal to generate a complementary drive waveform pair, and perform an orthogonal test based on the complementary drive waveform pair to establish a characteristic correlation matrix between the drive parameters and the display effect;

[0031] Specifically, edge detection is performed on the original PWM drive signal to accurately determine the rising-edge time sequence and falling-edge time sequence of the signal, and the duty cycle data of the drive signal is calculated based on this timing information. Since the duty cycle directly affects the gray-scale performance, brightness stability, and power consumption of the LED, during the PWM signal optimization process, it is necessary to accurately measure the duty cycle data and ensure the dynamic adjustment ability under different driving conditions. The original PWM drive signal is segmented based on the duty cycle data to achieve more refined drive control. The PWM signal is divided into a pre-charge segment waveform, a main drive segment waveform, and a stable segment waveform, and the timing synchronization between these three waveform segments is ensured. The pre-charge segment is used to adjust the transient response of the LED, the main drive segment is used to achieve the main light-emitting control, and the stable segment is used to maintain the stability of the signal to avoid a decrease in the display effect due to fluctuations in the power supply or drive characteristics. Through reasonable timing control, the current response of the LED pixel at different drive stages is optimized to improve the display uniformity and dynamic contrast. When performing PWM signal compensation, amplitude modulation is performed on the pre-charge segment waveform to generate a positive compensation pulse in the line scan signal to improve the response characteristics of the LED at the moment of lighting. An inverted negative compensation pulse is generated based on the waveform parameters of the positive compensation pulse to form a complementary effect during signal transmission. The matching degree of the positive compensation pulse and the negative compensation pulse directly affects the light-emitting uniformity of the LED. Therefore, it is necessary to accurately control their amplitude, duration, and phase relationship so that they can effectively suppress the non-linear distortion of the drive waveform during current transmission. To ensure that the compensation signal can play its role correctly, a compensation signal transfer function is established based on the positive compensation pulse and the negative compensation pulse, and the Laplace transform is used to perform frequency-domain analysis on this transfer function to obtain the spectral characteristics of the compensation signal. Through spectral analysis, the main harmonic components of the compensation signal and the introduced high-frequency noise are identified. On this basis, a crosstalk suppression filter is constructed. This filter is used to weaken unnecessary high-frequency harmonic components and optimize the transmission characteristics of the compensation signal in the drive system to reduce color shift or brightness fluctuations caused by signal interference. Under the action of the filter, the positive compensation pulse and the negative compensation pulse are filtered to obtain a corrected pair of compensation pulses, and the corrected pair of compensation pulses are respectively superimposed on the pre-charge segments of the line scan signal and the column scan signal to generate a pair of complementary drive waveforms. The role of the pair of complementary drive waveforms is to provide corresponding current regulation at different stages of LED lighting, make the light-emitting characteristics of the LED more stable, and reduce the color shift phenomenon in the high gray-scale region. By optimizing the rising edge and falling edge of the drive waveform, the electromagnetic interference that may be generated during the driving process is reduced, and the overall display performance of the LED array is improved. An orthogonal test is performed on the pair of complementary drive waveforms within a preset voltage range and a preset time range to systematically analyze the influence of different drive parameters on the display effect. Orthogonal testing is an experimental design method that can effectively evaluate the combined influence of multiple parameters with fewer test times.By setting factors such as different PWM duty cycles, compensation signal amplitudes, and filtering parameters, a series of orthogonal test experiments are carried out, and key indicators such as the brightness uniformity, gray-scale stability, and chromaticity consistency of the LED array are measured and analyzed. Based on the experimental data of the orthogonal test, a characteristic correlation matrix between the driving parameters and the display effect is established, and this matrix can quantify the display performance under different driving conditions. By analyzing the characteristic correlation matrix, the key driving parameters affecting the display effect are found, and the compensation strategy of the PWM signal is adjusted according to different display requirements, so as to achieve the best driving effect of the LED array.

[0032] Model the amplitude and pulse width of the compensation pulse and use them as orthogonal test factors to construct a systematic test plan. Since the amplitude of the compensation pulse has a direct impact on the brightness and response speed of the LED pixel, a reasonable division is made within the target voltage range, and the amplitude interval is divided into N test points to ensure that the test covers the entire driving range. At the same time, the pulse width determines the duration of the signal and the energy transfer characteristics, so it also needs to be divided into M test points within the target time range to obtain N×M groups of test parameters. Generate an orthogonal test table based on the N×M groups of parameters and use this test table to set the parameters of the complementary drive waveform pair to generate the corresponding experimental waveforms. The role of the orthogonal test table is to reduce the number of experiments while ensuring that all combinations of key parameters can be effectively covered. Therefore, uniformity and representativeness must be considered in the design. By setting the amplitude and pulse width of the compensation pulse of the PWM signal according to the orthogonal test table, N×M groups of experimental waveforms are obtained, and each group of experimental waveforms represents a possible configuration of the drive signal. During the experiment, waveform analysis is performed on key dynamic characteristic parameters such as the current rising edge change rate, overshoot amplitude, and crosstalk current of each group of experimental waveforms to quantify the electrical characteristics of each group of waveforms. The current rising edge change rate determines the response speed of the LED and directly affects the stability of high gray-scale display. Therefore, a high-speed oscilloscope or current sensor is needed to record the change of the rising edge during the test. At the same time, the measurement of the overshoot amplitude is crucial because the overshoot phenomenon will cause the instability of the drive signal, which in turn leads to uneven brightness or chromaticity shift. To ensure the display quality of the LED array, measure the magnitude of the crosstalk current to evaluate the impact of different waveform configurations on adjacent pixels, and find ways to reduce crosstalk through data analysis. After completing the measurement of the dynamic characteristic parameters, construct a multi-dimensional characteristic space based on these parameters and the aforementioned display quality evaluation function, so that all test data can be analyzed within a unified mathematical framework. To extract the main influencing factors in the data, perform a principal vector decomposition operation on this multi-dimensional characteristic space to obtain a set of eigenvectors. The main role of the principal vector decomposition is to extract the main change patterns in the data and remove redundant information to improve the efficiency and accuracy of the optimization calculation. Establish a mapping relationship between the drive parameters and the display effect according to the set of eigenvectors to form an optimization model. Use the least squares fitting calculation to optimize the parameters of the model and reduce the noise interference in the experimental data. The result of the least squares fitting generates a correlation coefficient matrix, which is used to describe the specific impact of different drive parameters on the display effect. Perform a singular value decomposition operation on the correlation coefficient matrix to remove redundant information and ensure that the mapping relationship between the drive parameters and the display effect has high interpretability and stability. Generate a characteristic correlation matrix between the drive parameters and the display effect.

[0033] Step S400: Calculate the optimal driving parameter combination according to the characteristic correlation matrix and the display quality evaluation function, and generate an LED array driving signal according to the optimal driving parameter combination.

[0034] Specifically, a target optimization function is constructed based on the characteristic correlation matrix to ensure that the optimization process can accurately reflect the display quality requirements of the LED array. Since the optimization of the driving waveform involves multiple interrelated parameters, when constructing the optimization objective, factors such as brightness uniformity, waveform distortion rate, and optical crosstalk suppression effect are comprehensively considered, and a unified optimization objective function is formed through a weighting strategy. To ensure the feasibility of the solution, constraint conditions are set for the target optimization function, and the amplitude range and pulse width range of the compensation pulse are used as the boundaries of the particle position, thereby limiting the target search space and enabling the optimization search to be carried out within a reasonable range to avoid solutions that exceed physical constraints. Particle initialization is performed on the target search space to generate an initial particle swarm, and the initial state of the particle swarm is optimized. Since conventional random initialization leads to uneven particle distribution, which in turn affects the optimization efficiency, during the particle initialization process, a chaotic sequence is constructed to perturb the initial particle swarm to enhance the diversity of the particle swarm and avoid the optimization process falling into a local optimal solution. Through this process, an initial solution space with a good search distribution is formed, and the fitness values of each initial particle are calculated based on the display quality evaluation function to evaluate the advantages and disadvantages of different initial solutions. To improve the stability and convergence speed of the optimization process, an adaptive inertia weight is calculated for the particles in the initial solution space. The role of the inertia weight is to adjust the movement trend of the particles in the search space, and the adaptive inertia weight dynamically adjusts its weight according to the fitness value of the particles, thereby maintaining a strong global search ability in the initial stage of optimization and gradually converging to local search in the later stage of optimization. This mechanism can effectively improve the convergence accuracy of the optimization algorithm and reduce the ineffective iterations in the search process. After calculating the inertia weight of the particle swarm, a velocity update operation is performed on the particle swarm based on the motion parameters, and a local search operator is introduced to adjust the particle position to form the candidate solution set of the next generation. The core of the velocity update operation is to adjust the moving direction of the particles according to the historical optimal solution of the particles and the global optimal solution of the population, so that it gradually converges to the optimal solution region. At the same time, the local search operator can enhance the local exploration ability of the optimization algorithm to ensure that the optimization result can be refined and adjusted within the local optimal region while meeting the global search requirements, thereby obtaining a better solution set. After generating the candidate solution set, a constraint check is performed on it to ensure that all solutions meet the physical constraint conditions. For particles that exceed the search boundary, the Lagrange multiplier method is used to process them to adjust them back to the feasible solution space and ensure that no invalid solutions are generated during the optimization process. The solution set after constraint processing is defined as the effective solution set, and it is sorted based on the display quality evaluation function to select the optimal solution. From the sorted effective solution set, the optimal particle position parameters are extracted, and these parameters are decoded and converted into compensation pulse parameters to form the optimal driving parameter combination. These optimal parameters can maximize the brightness uniformity of the LED array, reduce waveform distortion, and reduce optical crosstalk, thereby improving the overall performance of the LED display system.To ensure that the optimal drive parameters can be accurately used for the generation of PWM signals, a high-precision clock signal is generated by a digital phase-locked loop and input into a digital pulse width modulator to generate an LED array drive signal that meets the optimization requirements.

[0035] Digitally encode the optimal drive parameter combination so that it can adapt to the frequency synthesis and phase control mechanisms of a digital phase-locked loop (PLL). Convert the amplitude parameter of the compensation pulse into a frequency division coefficient to control the output frequency of the PLL. At the same time, convert the pulse width parameter into a phase control word to adjust the phase accumulation process of the PLL, obtaining the complete digital phase-locked loop control parameters. Construct a phase accumulator based on the digital phase-locked loop control parameters to ensure that the PLL can accurately generate a clock signal. The core function of the phase accumulator is to continuously accumulate the input phase increment and convert it into a sine signal through a numerically controlled oscillator. Therefore, during the phase accumulation process, perform a sine mapping operation on the accumulated output to obtain the reference signal of the phase frequency detector. Compare the reference signal with the feedback signal in terms of phase and frequency to evaluate the phase-locked state of the PLL. The phase frequency comparator calculates the phase error between the reference signal and the feedback signal and generates an error signal, which will be input into the digital loop filter for filtering. The role of the loop filter is to smooth the phase error, reduce high-frequency noise, and output a control voltage, which will be used as the input signal of the numerically controlled oscillator to accurately adjust the output frequency of the oscillator, thereby ensuring that the PLL can operate stably and generate a clock signal that meets the requirements of LED array driving. In the PLL system, the numerically controlled oscillator is the core component for generating the reference clock signal. It adjusts the oscillation frequency according to the control voltage of the loop filter to ensure that the output signal is consistent with the target frequency. To improve the stability and flexibility of the clock signal, perform frequency division on the reference clock signal to obtain the feedback clock signal, thereby forming a complete phase-locked control loop in the PLL closed-loop system. The role of frequency division is to adjust the period of the clock signal to adapt to different LED driving requirements and ensure that the final PWM signal can meet the response characteristics and display quality requirements of the LEDs. After generating a stable reference clock signal, establish a counter for the digital pulse width modulation unit (DPWM) based on this signal to ensure the accurate control of the PWM signal. In the DPWM module, the role of the counter is to continuously accumulate the reference clock signal and determine the duty cycle of the PWM signal according to the count value. Therefore, during the counting process, compare the count value with the duty cycle data of the compensation pulse to generate a pulse width modulation waveform. The accuracy and stability of this pulse width modulation waveform determine the grayscale display effect of the LED pixel points and directly affect the contrast and dynamic response ability of the display screen. To ensure that the PWM signal can drive the LED array, perform level conversion and current amplification on the PWM signal through a driver circuit to match the working voltage and current requirements of the LEDs. The driver circuit consists of a power MOSFET or a constant current driving chip, which can convert the low-power PWM control signal into a high-current driving signal suitable for the LED array and ensure the linearity of the driving current to reduce color deviation and brightness unevenness in the LED display screen.Through digital signal processing and hardware optimization, a stable and efficient LED array driving signal is generated, enabling the small-pitch LED display screen to achieve the best state in terms of brightness uniformity, response speed, and color accuracy, thereby enhancing the overall display effect and system reliability.

[0036] In the embodiments of the present invention, by establishing a complete mathematical model of the light-emitting characteristics of pixel points and combining temperature characteristic compensation, an accurate description of the display characteristics of the LED array is achieved. Adopting a design scheme of complementary driving waveform pairs, by injecting positive and negative compensation pulses into the row and column scanning signals, the voltage spikes and crosstalk effects during the driving process are effectively suppressed, and the uniformity of the display screen is improved. An establishment method of a characteristic correlation matrix based on orthogonal testing is introduced. Through systematic parameter testing and data analysis, an accurate mapping relationship between driving parameters and display effects is established. An improved particle swarm optimization algorithm is designed. By initializing with a chaotic sequence and adjusting the adaptive weight, the convergence efficiency and the quality of the solution in the optimization process are improved. A driving signal generation scheme combining a digital phase-locked loop and digital pulse width modulation is adopted to achieve high-precision control of the driving waveform and ensure the stability and reliability of the driving signal.

[0037] In a specific embodiment, the process of executing step S100 may specifically include the following steps:

[0038] Detect the ambient temperature of the small-pitch LED array to obtain temperature acquisition data, and divide the current sampling interval based on the temperature acquisition data;

[0039] Input the driving current into each pixel point in the small-pitch LED array in sequence according to the current sampling interval, and collect the spectral data of each pixel point through a spectral analyzer to obtain spectral wavelength data and spectral intensity data;

[0040] Perform a dominant wavelength extraction operation on the spectral wavelength data to obtain current-wavelength characteristic data, and perform an integration operation on the spectral intensity data to obtain current-brightness characteristic data;

[0041] Based on the current-wavelength characteristic data, establish a wavelength characteristic polynomial function, and perform a least squares fitting operation on the wavelength characteristic polynomial function to obtain a wavelength characteristic coefficient matrix;

[0042] Based on the current-brightness characteristic data, establish a brightness characteristic polynomial function, and perform a least squares fitting operation on the brightness characteristic polynomial function to obtain a brightness characteristic coefficient matrix;

[0043] Based on the wavelength characteristic coefficient matrix and the brightness characteristic coefficient matrix, establish a three-dimensional characteristic surface to generate a mathematical model of the light-emitting characteristics of pixel points.

[0044] Specifically, a high-precision temperature sensor is used to monitor the operating ambient temperature of the LED display in real time. The light-emitting characteristics of LEDs, including spectral wavelength and brightness response, shift with temperature changes, and this shift directly affects the color consistency and brightness uniformity of the display. Assume the data obtained from temperature measurement is ( ), where represents the temperature values of different pixel points at different times. By statistically analyzing these temperature data, determine the range of temperature change , and based on this range, divide the current sampling intervals so that the current input can match the light-emitting characteristics of the LEDs at different temperatures. Input different driving currents ( ) into each pixel point of the small-pitch LED array in sequence, and use a spectral analyzer to collect the corresponding spectral data, including spectral wavelength data and spectral intensity data , where represents the main emission wavelength at temperature and driving current , while represents the spectral intensity distribution under the corresponding conditions. After obtaining the spectral wavelength data, perform the main wavelength extraction operation, and use the spectral centroid method to calculate the main emission wavelength:

[0045]

[0046] where, is the spectral intensity value at a specific wavelength . This calculation formula can effectively extract the main emission wavelength of the LED and reflect the influence of driving current and temperature on spectral drift. At the same time, in order to obtain the brightness characteristic data of the LED, perform an integration operation on the spectral intensity data to calculate the total luminous flux:

[0047]

[0048] where, represents the brightness of the LED at temperature and driving current . This integration calculation can accurately describe the change of the LED's luminous intensity with current and temperature and is used for subsequent brightness compensation optimization. After obtaining the current-wavelength characteristic data , establish a wavelength characteristic polynomial function to fit the data relationship, using a quadratic or cubic polynomial:

[0049]

[0050] where, is the coefficient to be determined, representing the influence of current and temperature on wavelength. To determine these coefficients, a least-squares fitting operation is performed on this polynomial, and its goal is to minimize the error function:

[0051]

[0052] The optimal solution can be obtained through matrix solution method, and the wavelength characteristic coefficient matrix is obtained:

[0053]

[0054] Similarly, for the luminance characteristic data , a luminance characteristic polynomial function is established:

[0055]

[0056] where are the luminance characteristic fitting coefficients, which are optimized by the least-squares method to solve the error function:

[0057]

[0058] Finally, the luminance characteristic coefficient matrix is obtained:

[0059] ;

[0060] Based on the wavelength characteristic coefficient matrix and the luminance characteristic coefficient matrix a three-dimensional characteristic surface is constructed. This surface takes the drive current and the temperature as independent variables, and the wavelength and the luminance as response variables to form a mathematical model of the pixel point light emission characteristics:

[0061] .

[0062] In a specific embodiment, the process of executing step S200 may specifically include the following steps:

[0063] Perform a numerical integration operation on the mathematical model of the pixel point light emission characteristics to obtain the average luminance data of each pixel point, and calculate the ratio of the maximum luminance value to the minimum luminance value according to the average luminance data to obtain the luminance uniformity coefficient;

[0064] Perform a Fourier transform on the waveform response parameters in the mathematical model of the pixel point light emission characteristics to obtain the spectral component data, and calculate the harmonic distortion factor according to the spectral component data to obtain the waveform distortion rate;

[0065] An optical crosstalk model between pixel points is established according to the mathematical model of pixel point light emission characteristics. Finite element analysis is performed on the optical crosstalk model to obtain the crosstalk attenuation ratio.

[0066] A weight matrix is constructed based on the brightness uniformity coefficient, waveform distortion rate, and crosstalk attenuation ratio, and eigenvalue decomposition operation is performed on the weight matrix to obtain weight coefficients.

[0067] The brightness uniformity coefficient, waveform distortion rate, and crosstalk attenuation ratio are normalized to obtain standardized evaluation indicators, and the standardized evaluation indicators are linearly weighted according to the weight coefficients to obtain the initial evaluation score.

[0068] The initial evaluation score is transformed through the hyperbolic tangent function to obtain the display quality evaluation function.

[0069] Specifically, based on the mathematical model of pixel point light emission characteristics The brightness function expressed , that is, the brightness distribution of the pixel point at the coordinate under different driving currents and ambient temperature . To calculate the average brightness of each pixel point, numerical integration is performed on the driving time of the entire pixel point:

[0070]

[0071] Among them, represents the average brightness of the pixel point within the driving period , while represents the brightness of the pixel point changing with time. To measure the overall brightness uniformity of the LED array, the maximum brightness value and the minimum brightness value in the pixel point array are found and the brightness uniformity coefficient is calculated:

[0072]

[0073] Among them, The value range is between (0, 1]. When approaches 1, it indicates that the brightness uniformity of the entire LED array is better, while when is smaller, it indicates that there is a large brightness non-uniformity phenomenon. To analyze the waveform response characteristics of the pixel point, Fourier transform is performed on the waveform response parameters in the mathematical model of pixel point light emission characteristics to obtain the spectral component data of the signal. Assuming that the PWM driving signal of the pixel point is , then its Fourier transform is represented by the following formula:

[0074]

[0075] Among them, is the frequency component, representing the amplitude distribution of the driving signal at different frequencies. To measure the waveform distortion degree of the signal, it is necessary to calculate the harmonic distortion factor , and its definition is:

[0076]

[0077] Among them, represents the th harmonic component, represents the fundamental wave component. When is small, it means that the distortion of the PWM signal is small, so the light output of the LED is more stable; on the contrary, a larger will cause problems such as display screen flickering and gray scale instability. When analyzing the optical crosstalk problem of the LED array, an optical crosstalk model is established based on the mathematical model of the light emission characteristics of pixel points. The optical crosstalk mainly comes from the light radiation diffusion between adjacent pixel points, so the finite element analysis method is used to simulate the optical energy transmission between pixel points. Assuming that the optical power distribution of adjacent pixel points is , the basic equation of the optical crosstalk model is:

[0078]

[0079] Among them, is the light wave propagation constant, represents the Laplace operator. Through finite element solution, the crosstalk attenuation ratio is obtained:

[0080] ;

[0081] Among them, represents the optical power of the original pixel point. When is small, it indicates that the influence of optical crosstalk is small, while a larger may cause color non-uniformity of the LED display screen. Based on the brightness uniformity coefficient , the waveform distortion rate and the crosstalk attenuation ratio , a weight matrix is constructed to characterize the influence of each parameter on the display quality. Assuming that the weight matrix is:

[0082]

[0083] Through eigenvalue decomposition operation, the weight coefficient is solved:

[0084] ;

[0085] Among them, is the eigenvector, is the corresponding weight coefficient. The weight coefficient calculated based on the eigenvalue is used for subsequent quality evaluation calculations. In order to standardize various evaluation indicators, the luminance uniformity coefficient, waveform distortion rate, and crosstalk attenuation ratio are normalized, and the normalization variable is defined as:

[0086]

[0087] Among them, represents the original data, and represent the minimum and maximum values of the data respectively. After normalization, the weight coefficient is used to linearly weight the standardized evaluation indicators to obtain the initial evaluation score :

[0088] ;

[0089] The initial evaluation score is non-linearly transformed through the hyperbolic tangent function to obtain the final display quality evaluation function :

[0090] ;

[0091] Among them, serves to compress the evaluation value into the range of (-1, 1) and enhance the discrimination ability for extreme values.

[0092] In a specific embodiment, the process of executing step S300 may specifically include the following steps:

[0093] Perform edge detection on the original PWM drive signal to obtain the rising edge time sequence and the falling edge time sequence, and calculate the duty cycle data of the drive signal based on the rising edge time sequence and the falling edge time sequence;

[0094] Based on the duty cycle data, segment the original PWM drive signal to obtain the pre-charge segment waveform, the main drive segment waveform, and the stable segment waveform, and perform timing synchronization on the pre-charge segment waveform, the main drive segment waveform, and the stable segment waveform;

[0095] Perform amplitude modulation on the pre-charge segment waveform to generate a positive compensation pulse in the line scan signal, and generate an inverted negative compensation pulse according to the waveform parameters of the positive compensation pulse;

[0096] A compensation signal transfer function is established based on positive compensation pulses and negative compensation pulses, and the frequency-domain analysis of the compensation signal transfer function is performed through Laplace transform to obtain the compensation signal spectrum.

[0097] A crosstalk suppression filter is constructed according to the compensation signal spectrum, and the positive compensation pulse and negative compensation pulse are filtered to obtain a pair of corrected compensation pulses, and the pair of corrected compensation pulses are respectively superimposed on the pre-charge sections of the row scan signal and the column scan signal to generate a pair of complementary drive waveforms.

[0098] An orthogonal test is performed on the pair of complementary drive waveforms within a preset voltage range and a preset time range to establish a characteristic correlation matrix between drive parameters and display effects.

[0099] Specifically, time-domain sampling is performed on the PWM signal to obtain the rising-edge time sequence and the falling-edge time sequence. Let the period of the PWM signal be , and its timing function is expressed as:

[0100] ;

[0101] Among them, and represent the rising-edge and falling-edge time sequences respectively, is the high level of the PWM signal, is the low level. In order to calculate the duty cycle data of the PWM signal, calculate the high-level duration in each period and normalize it to obtain the duty cycle :

[0102]

[0103] This duty cycle data is used for further optimization of the PWM signal to match the response characteristics of the LED pixel points. After obtaining the duty cycle of the PWM signal, the signal is segmented to obtain the pre-charge section waveform, the main drive section waveform, and the stable section waveform. These three stages correspond to the pre-charge, main drive process, and signal stable stage of the LED respectively, so they need to be synchronized in timing to ensure a smooth transition of the current input in each stage. The segmented waveform is expressed as:

[0104] );

[0105] Among them, , and are the durations of the pre-charge, main drive, and stable stages respectively, are the voltage amplitudes of these three stages respectively. In order to optimize the response characteristics of the LED, amplitude modulation is introduced in the pre-charge stage to generate a positive compensation pulse in the row scan signal:

[0106] ;

[0107] Among them, is the amplitude of the compensation pulse, is the time constant, is the duration of the compensation pulse. Since the compensation pulse causes additional signal disturbance, an inverted negative compensation pulse is generated to cancel the corresponding interference:

[0108] ;

[0109] These two complementary pulses are used to adjust the driving response of the LED and reduce high-frequency noise. To analyze the impact of the compensation pulse on the system, the transfer function of the compensation signal is established and Laplace-transformed to represent it in the frequency domain from the time domain:

[0110]

[0111] Among them, is the Laplace-transform variable. By calculating the spectral response of the compensation signal:

[0112] ;

[0113] The amplitudes of different frequency components are obtained, and a crosstalk suppression filter is constructed to eliminate high-frequency noise. The transfer function of the crosstalk suppression filter is expressed as:

[0114]

[0115] Among them, is the cut-off frequency of the filter. The filtered compensation pulse pair is:

[0116] ;

[0117] This signal is used to adjust the row scanning and column scanning pre-charge segments of the LED array, thereby generating a pair of complementary drive waveforms:

[0118]

[0119] ;

[0120] After generating the pair of complementary drive waveforms, an orthogonal test is performed within the preset voltage range and the preset time range to analyze the influence of the drive parameters on the display effect. The key to the orthogonal test lies in constructing a test matrix, taking the amplitude of the compensation pulse and the time constant as test factors, and defining test points:

[0121]

[0122]

[0123] Among them, , . Orthogonal testing generates sets of test parameters, and each set of parameters is used to drive the experiment and measure the brightness uniformity, gray-scale response, and crosstalk intensity of the LED. The experimental data is organized into a characteristic correlation matrix:

[0124] ;

[0125] Among them, represents the evaluation value of the display effect of each set of driving parameters. By calculating the optimal parameter set, the driving characteristics of the PWM signal are optimized, and the brightness uniformity and visual effect of the LED display are improved.

[0126] Among them, after generating the complementary driving waveform pair and before performing the orthogonal test, it further includes: performing time-domain discretization processing on the complementary driving waveform pair, dividing the driving period into N preset time segments to obtain a discrete sampling sequence; marking the key waveform feature points in the discrete sampling sequence, dividing the driving waveform into a steady-state interval and a transient interval to obtain partitioned sampling data; constructing an adaptive integration model based on the partitioned sampling data, using a variable-step integration algorithm for the transient interval and a fixed-step integration algorithm for the steady-state interval to obtain mixed integration parameters; numerically reconstructing the complementary driving waveform pair according to the mixed integration parameters, using a high-order integration algorithm to perform fine-grained calculation on the transient interval to obtain optimized waveform data; establishing an error evaluation function based on the optimized waveform data, obtaining the integration error of each interval through recursive calculation, and comparing it with a preset error threshold to obtain an error distribution matrix; adaptively adjusting the integration step according to the error distribution matrix, refining the calculation step in the area where the error exceeds the threshold and keeping the step unchanged in the area where the error meets the standard to obtain corrected integration parameters; iteratively optimizing the complementary driving waveform pair using the corrected integration parameters until the integration error of all intervals is less than the preset threshold to obtain the final driving waveform data; smoothing the final driving waveform data through a digital filter to generate an optimized complementary driving waveform pair.

[0127] In a specific embodiment, the process of performing the steps of performing an orthogonal test on the complementary driving waveform pair within a preset voltage range and a preset time range and establishing a characteristic correlation matrix between the driving parameters and the display effect may specifically include the following steps:

[0128] Construct orthogonal test factors for the compensation pulse amplitude and pulse width of the complementary drive waveform pair. Divide the compensation pulse amplitude into N test points within the target voltage range, and divide the pulse width into M test points within the target time range to obtain N×M sets of test parameters;

[0129] Generate an orthogonal test table based on the N×M sets of test parameters, and set the parameters of the complementary drive waveform pair according to the orthogonal test table to obtain N×M sets of experimental waveforms;

[0130] Perform waveform analysis on the current rise rate, overshoot amplitude, and crosstalk current of the N×M sets of experimental waveforms to obtain the dynamic characteristic parameters of each set of waveforms;

[0131] Construct a multi-dimensional characteristic space based on the dynamic characteristic parameters and the display quality evaluation function, and perform principal vector decomposition operation on the multi-dimensional characteristic space to obtain a set of eigenvectors;

[0132] Establish the mapping relationship between the drive parameters and the display effect according to the set of eigenvectors, perform the least squares fitting calculation on the mapping relationship to obtain the correlation coefficient matrix, and perform the singular value decomposition operation on the correlation coefficient matrix to generate the characteristic correlation matrix between the drive parameters and the display effect.

[0133] Specifically, construct orthogonal test factors for the compensation pulse amplitude and pulse width of the complementary drive waveform pair, and determine the voltage amplitude range and pulse width range of the compensation pulse to ensure the coverage of the test. Set the target range of the compensation pulse amplitude as , and divide it into test points. The amplitude of each test point is expressed as:

[0134]

[0135] where represents the compensation pulse amplitude of the th test point. Similarly, set the target width range of the compensation pulse as , and divide it into test points. The pulse width of each test point is:

[0136]

[0137] where represents the compensation pulse width of the th test point. Through the combination of the amplitude and width of the compensation pulse, sets of test parameters are constructed to form a complete experimental test matrix. Based on Generate an orthogonal test table with a set of parameters, and use this orthogonal test table to set the parameters for complementary drive waveform pairs to generate corresponding experimental waveforms. Orthogonal testing is an optimization experimental method that can obtain the optimal parameter combination within a limited number of tests. The orthogonal test table is represented by the following matrix:

[0138] ;

[0139] Each row represents a set of test parameters used to control the amplitude and pulse width of the complementary drive waveform, enabling the experimental tests to be evenly distributed across the entire parameter space. Analyze the waveforms of the set of experimental waveforms to evaluate their dynamic characteristics. Measure the rate of change of the current rising edge , overshoot amplitude and crosstalk current of the experimental waveforms. Given the instantaneous current of the waveform, the rate of change of the rising edge is expressed as:

[0140]

[0141] where is the time point of the rising edge of the waveform. The overshoot amplitude is the maximum amplitude exceeding the steady-state current value in the transient response of the signal, defined as:

[0142] ;

[0143] where is the current value after the signal stabilizes. The crosstalk current is the current component caused by interference between adjacent pixel points, and its calculation formula is:

[0144]

[0145] where is the parasitic current of the adjacent pixel point, is the current of the main drive pixel point. After obtaining the dynamic characteristic parameters of all experimental waveforms, construct a multi-dimensional characteristic space based on these parameters and the display quality evaluation function. Define each data point in the characteristic space as:

[0146] ;

[0147] The matrix representation of the entire characteristic space is:

[0148] ;

[0149] To reduce the dimension of the data and extract the main features, a principal vector decomposition operation is performed on this feature space, that is, on perform eigenvalue decomposition:

[0150] ;

[0151] where is the eigenvector, is the corresponding eigenvalue. By calculating the eigenvector corresponding to the largest eigenvalue, the main influencing factors are obtained and used to establish the mapping relationship between the driving parameters and the display effect. Perform a least squares fitting calculation on the mapping relationship to obtain the correlation coefficient matrix . Set the driving parameter matrix and the display effect matrix :

[0152] ;

[0153] where is the display quality evaluation value under each set of driving parameters. Then the goal of the least squares fitting calculation is to solve such that:

[0154] ;

[0155] Calculate by the least squares method:

[0156] ;

[0157] Finally, the correlation coefficient matrix is obtained, which is used to describe the quantitative relationship between the driving parameters and the display effect. Perform singular value decomposition on the correlation coefficient matrix to eliminate redundant information and improve the stability of the calculation. The mathematical expression of singular value decomposition is:

[0158]

[0159] where and are the left and right singular vector matrices respectively, is a diagonal matrix, and its diagonal elements are singular values. By screening the larger singular values, the influence of noise is removed, and the final characteristic correlation matrix of the driving parameters and the display effect is generated:

[0160] .

[0161] Among them, after generating the characteristic correlation matrix of the driving parameters and the display effects and before calculating the optimal combination of driving parameters, it further includes: performing eigenanalysis on the characteristic correlation matrix, extracting the main influencing factors, establishing an objective function for parameter optimization, and determining the optimization constraint conditions according to the physical characteristics of the LED array to obtain an initial optimization model; constructing a bivariate iterative structure based on the initial optimization model, mapping the amplitude optimization and timing optimization of the driving parameters to the original variable space and the dual variable space respectively to obtain a solution space mapping matrix; setting a dynamic trigger threshold for the solution space mapping matrix, establishing an adaptive sampling strategy according to the parameter sensitivity to obtain a parameter update trigger condition; performing a strong ball search based on the parameter update trigger condition, constructing a search trajectory by calculating the gradient direction and step size of each point in the solution space to obtain a candidate solution sequence; performing convergence analysis on the candidate solution sequence, establishing an iterative termination criterion, and calculating the convergence speed of the optimization process according to the criterion to obtain a convergence characteristic curve; dynamically adjusting the search step size according to the convergence characteristic curve, increasing the step size in the fast convergence region and decreasing the step size in the fine search region to obtain an adaptive step size sequence; updating the parameters of the strong ball search process based on the adaptive step size sequence, and determining whether to execute the update operation by judging the trigger condition to obtain an optimized iteration sequence; inputting the optimized iteration sequence into the primal-dual iterator to perform alternating optimization of the parameter space until the convergence condition is satisfied to generate a preprocessed characteristic correlation matrix.

[0162] In a specific embodiment, the process of executing step S400 may specifically include the following steps:

[0163] Constructing an objective optimization function based on the characteristic correlation matrix, setting constraint conditions for the objective optimization function, and taking the compensation pulse amplitude range and pulse width range as the particle position boundaries to obtain an objective search space;

[0164] Initializing particles for the objective search space, constructing a chaotic sequence to perturb the initial particle swarm to obtain an initial solution space, and calculating the fitness values of each initial particle based on the display quality evaluation function;

[0165] Calculating the adaptive inertia weight for the particles in the initial solution space, dynamically adjusting the inertia weight according to the fitness values to obtain the motion parameters of the particle swarm;

[0166] Performing a velocity update operation on the particle swarm based on the motion parameters, and introducing a local search operator to update the particle positions to obtain an iterated candidate solution set;

[0167] Performing constraint checking on the candidate solution set, processing the out-of-bounds particles by the Lagrange multiplier method to obtain an effective solution set, and sorting the effective solution set according to the display quality evaluation function;

[0168] Extract the optimal particle position parameters from the effective solution set, decode and convert the optimal particle position parameters into compensation pulse parameters, and generate an optimal drive parameter combination;

[0169] Generate a clock signal from the optimal drive parameter combination through a digital phase-locked loop, and input the clock signal into a digital pulse width modulator to generate an LED array drive signal.

[0170] Specifically, use the characteristic correlation matrix Establish an objective optimization function to characterize the impact of drive parameters on display quality. Let be the optimization variable, where is the compensation pulse amplitude, is the compensation pulse width, then the optimization objective is defined as maximizing the display quality evaluation function :

[0171] ;

[0172] To make the optimization search space reasonable, impose constraint conditions on the objective optimization function. Let the range of the compensation pulse amplitude be , and the range of the pulse width be , then the search space is defined as:

[0173] ;

[0174] After establishing the search space, use the particle swarm optimization algorithm for optimization and solution. Initialize the particles for the target search space, and set the particle swarm size to , and the initial position of each particle is randomly distributed within , that is:

[0175] ;

[0176] Among them, represents a uniform distribution within the interval . To improve the diversity of the initial particles, use a chaotic sequence to perturb the particle swarm. Use the Logistic chaotic map:

[0177] ;

[0178] Among them, is the chaotic parameter (usually taken as 3.99), is the chaotic sequence value, and this sequence is used to adjust the distribution of the initial particles, so as to avoid early convergence to the local optimal solution. After initializing the particle swarm, calculate the fitness value of each particle, that is, the display quality evaluation function :

[0179]

[0180] Based on the fitness value, an adaptive inertia weight calculation is performed on the particle swarm to dynamically adjust the search behavior of the particles. The inertia weight Adopts the following adaptive formula:

[0181]

[0182] Where and Are the maximum and minimum values of the inertia weight respectively, and Are the maximum and minimum fitness values in the current particle swarm. The adaptive inertia weight can improve the convergence speed, and at the same time ensure global search in the early stage and local optimization in the later stage. After updating the inertia weight, a velocity update operation is performed on the particle swarm, and the velocity of particle Is calculated by the following formula:

[0183]

[0184] Where Is the acceleration factor, Is a random number, Is the historical best position of particle , Is the global best position. The position of the particle is updated as:

[0185]

[0186] To enhance the local search ability, a local search operator is introduced after each iteration to fine-tune the candidate solution set. Gaussian perturbation is used:

[0187]

[0188] Where Represents Gaussian noise with a mean of 0 and a variance of . A constraint test is performed on the candidate solution set to ensure that all particles are located within the legal search space . For out-of-bounds particles, the Lagrange multiplier method is used for correction. Let the out-of-bounds objective function be:

[0189] ;

[0190] Solve So that Satisfies the boundary conditions to obtain the corrected effective solution set. In the effective solution set, the particles are sorted according to the fitness value, and the optimal particle position As the final compensation pulse parameters, decode these parameters into the optimal drive parameter combination and use them for drive signal generation. Generate a clock signal through a digital phase-locked loop (PLL), where the input clock Through a frequency divider Generate the output clock :

[0191]

[0192] Among them, the frequency division coefficient Is calculated from the compensation pulse width Input the clock signal into a digital pulse width modulator (PWM), where the pulse width Is used to set the duty cycle :

[0193]

[0194] The PWM signal is converted into an LED array drive signal through a drive circuit :

[0195] ;

[0196] Through the above optimization process, adaptively adjust the amplitude and width of the compensation pulse to achieve the best LED display effect.

[0197] In a specific embodiment, the process of executing the steps of generating a clock signal through a digital phase-locked loop with the optimal drive parameter combination and inputting the clock signal into a digital pulse width modulator to generate an LED array drive signal may specifically include the following steps:

[0198] Perform digital encoding on the optimal drive parameter combination, convert the compensation pulse amplitude parameter into a frequency division coefficient, and convert the pulse width parameter into a phase control word to obtain the digital phase-locked loop control parameter;

[0199] Construct a phase accumulator based on the digital phase-locked loop control parameter, perform a sine mapping operation on the accumulated output of the phase accumulator to obtain the reference signal of the phase frequency detector;

[0200] Perform a phase frequency comparison on the reference signal and the feedback signal, and filter the comparison result through a digital loop filter to obtain the control voltage of the numerically controlled oscillator;

[0201] Input the control voltage into the numerically controlled oscillator for frequency modulation, generate a reference clock signal, and perform frequency division processing on the reference clock signal to obtain a feedback clock signal;

[0202] A counter of the digital pulse width modulation unit is established based on a reference clock signal, and a comparison operation is performed on the count value of the counter and the duty cycle data of the compensation pulse to obtain a pulse width modulation waveform. Then, the pulse width modulation waveform is subjected to level conversion and current amplification through a driver circuit to generate an LED array drive signal.

[0203] Specifically, digital encoding is performed on the optimal drive parameter combination to adapt to the control logic of the digital phase-locked loop. Let the optimal compensation pulse amplitude be , and the optimal pulse width be . Then, the compensation pulse amplitude is first converted into the frequency division coefficient of the PLL, while the pulse width parameter is converted into the phase control word . The calculation of the frequency division coefficient is based on the relationship between the target output frequency and the input reference frequency , satisfying:

[0204]

[0205] Wherein, must be an integer, and quantization is performed by finding the closest integer value. And the phase control word needs to match the pulse width . The phase step is controlled through a phase accumulator, and the calculation formula is:

[0206]

[0207] Wherein, is the clock period, is the bit width of the phase accumulator, such that has a value range of , and is used to control the phase increment. A phase accumulator is constructed based on the digital phase-locked loop control parameters. The function of this accumulator is to accumulate the phase increment and the current phase value to control the change of the phase. The calculation formula is:

[0208]

[0209] Wherein, represents the currently accumulated phase value, while controls the phase step. Since the output value of the phase accumulator is between , a sine mapping operation needs to be performed to obtain the reference signal of the phase frequency detector. This mapping is expressed as:

[0210]

[0211] Wherein, is the amplitude of the oscillation signal, ensuring that the reference signal is used for subsequent phase frequency comparison. The reference signal With the feedback signal Perform phase-frequency comparison. Calculate the phase error :

[0212] ;

[0213] This error signal enters a digital loop filter, which is used to smooth the phase error and remove high-frequency noise. The transfer function of the loop filter usually adopts a second-order IIR filter:

[0214]

[0215] Wherein, and are the proportional and integral gains respectively, controls the filtering time constant, and the filtered signal is used to control a numerically controlled oscillator to generate a modulated output frequency. The control voltage performs frequency modulation through a numerically controlled oscillator (DCO), and the output frequency of the DCO is determined by the control voltage:

[0216] ;

[0217] Wherein, is the center frequency of the DCO, is the gain factor of the DCO. The reference clock signal output by the DCO generates a feedback clock signal through a programmable divider:

[0218]

[0219] This feedback signal is used for the closed-loop control of the phase-locked loop to ensure the stable operation of the system. After obtaining a stable reference clock signal, a counter of a digital pulse width modulation (PWM) unit is established based on this signal to achieve the duty cycle control of the PWM signal. Assume the initial value of the counter is 0, and it increments in each clock cycle until it reaches the maximum value and then resets to zero:

[0220]

[0221] The value of the counter is compared with the duty cycle data of the compensation pulse to generate a PWM waveform:

[0222]

[0223] Wherein, is the high level of the PWM signal, which is used to control the LED drive current. The PWM waveform undergoes level conversion and current amplification through a drive circuit to generate a drive signal for the LED array. Assume the gain of the driver is , the output drive current is:

[0224] .

[0225] Please refer to Figure 2 , Figure 2 , which is a schematic block diagram of the small-pitch LED array drive waveform optimization device 200 provided by the embodiment of the present application. As Figure 2 shown, the small-pitch LED array drive waveform optimization device 200 includes:

[0226] An acquisition module 210, configured to collect spectral characteristics of pixel points in the small-pitch LED array and construct a mathematical model of the light-emitting characteristics of the pixel points;

[0227] A construction module 220, configured to perform parameter operations on the mathematical model of the light-emitting characteristics of the pixel points to obtain array display parameters, and construct a display quality evaluation function according to the array display parameters;

[0228] A test module 230, configured to perform waveform compensation on the original PWM drive signal to generate a complementary drive waveform pair, and perform an orthogonal test based on the complementary drive waveform pair to establish a characteristic correlation matrix between the drive parameters and the display effect;

[0229] A generation module 240, configured to calculate an optimal drive parameter combination according to the characteristic correlation matrix and the display quality evaluation function, and generate an LED array drive signal according to the optimal drive parameter combination.

[0230] Through the collaborative cooperation of the above-mentioned various components, by establishing a complete mathematical model of the light-emitting characteristics of pixel points and combining temperature characteristic compensation, an accurate description of the display characteristics of the LED array is achieved. The design scheme of the complementary drive waveform pair effectively suppresses voltage spikes and crosstalk effects during the driving process by injecting positive and negative compensation pulses into the row and column scanning signals, improving the uniformity of the display screen. An orthogonal test-based characteristic correlation matrix establishment method is introduced, and an accurate mapping relationship between the drive parameters and the display effect is established through systematic parameter testing and data analysis. An improved particle swarm optimization algorithm is designed, and the convergence efficiency and solution quality of the optimization process are improved through chaotic sequence initialization and adaptive weight adjustment. A drive signal generation scheme combining a digital phase-locked loop and digital pulse width modulation is adopted to achieve high-precision control of the drive waveform and ensure the stability and reliability of the drive signal.

[0231] Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the above-described systems, systems, and units can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0232] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of this application. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0233] As described above, the above embodiments are only used to illustrate the technical solutions of this application and are not intended to limit them; although this application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of various embodiments of this application.

Claims

1. A method for optimizing the driving waveform of a small-pitch LED array, characterized in that Including: Collecting the spectral characteristics of pixel points in a small-pitch LED array to construct a mathematical model of the luminous characteristics of pixel points; specifically including: detecting the ambient temperature of the small-pitch LED array to obtain temperature acquisition data, and dividing the current sampling interval based on the temperature acquisition data; sequentially inputting the driving current into each pixel point in the small-pitch LED array according to the current sampling interval, and collecting the spectral data of each pixel point through a spectral analyzer to obtain spectral wavelength data and spectral intensity data; performing a dominant wavelength extraction operation on the spectral wavelength data to obtain current-wavelength characteristic data, and performing an integration operation on the spectral intensity data to obtain current-brightness characteristic data; establishing a wavelength characteristic polynomial function based on the current-wavelength characteristic data, and performing a least-squares fitting operation on the wavelength characteristic polynomial function to obtain a wavelength characteristic coefficient matrix; establishing a brightness characteristic polynomial function based on the current-brightness characteristic data, and performing a least-squares fitting operation on the brightness characteristic polynomial function to obtain a brightness characteristic coefficient matrix; establishing a three-dimensional characteristic surface based on the wavelength characteristic coefficient matrix and the brightness characteristic coefficient matrix to generate a mathematical model of the luminous characteristics of pixel points; Performing parameter operations on the mathematical model of the luminous characteristics of pixel points to obtain array display parameters, and constructing a display quality evaluation function based on the array display parameters; Performing waveform compensation on the original PWM driving signal to generate a complementary driving waveform pair, and performing an orthogonal test based on the complementary driving waveform pair to establish a characteristic correlation matrix between driving parameters and display effects; Calculating an optimal driving parameter combination according to the characteristic correlation matrix and the display quality evaluation function, and generating an LED array driving signal according to the optimal driving parameter combination.

2. The method for optimizing the driving waveform of the small-pitch LED array according to claim 1, wherein The performing parameter operations on the mathematical model of the luminous characteristics of pixel points to obtain array display parameters, and constructing a display quality evaluation function based on the array display parameters includes: Performing a numerical integration operation on the mathematical model of the luminous characteristics of pixel points to obtain the brightness mean data of each pixel point, and calculating the ratio of the maximum brightness value to the minimum brightness value according to the brightness mean data to obtain a brightness uniformity coefficient; Performing a Fourier transform on the waveform response parameters in the mathematical model of the luminous characteristics of pixel points to obtain spectral component data, and calculating a harmonic distortion factor according to the spectral component data to obtain a waveform distortion rate; Establishing an optical crosstalk model between pixel points according to the mathematical model of the luminous characteristics of pixel points, and performing a finite element analysis on the optical crosstalk model to obtain a crosstalk attenuation ratio; Constructing a weight matrix based on the brightness uniformity coefficient, the waveform distortion rate, and the crosstalk attenuation ratio, and performing an eigenvalue decomposition operation on the weight matrix to obtain weight coefficients; Performing a normalization process on the brightness uniformity coefficient, the waveform distortion rate, and the crosstalk attenuation ratio to obtain standardized evaluation indicators, and performing a linear weighting on the standardized evaluation indicators according to the weight coefficients to obtain an initial evaluation score; Performing a transformation operation on the initial evaluation score through a hyperbolic tangent function to obtain a display quality evaluation function.

3. The method for optimizing the driving waveform of the small-pitch LED array according to claim 1, wherein Waveform compensation is performed on the original PWM drive signal to generate a complementary drive waveform pair, and an orthogonal test is performed based on the complementary drive waveform pair to establish a characteristic correlation matrix between drive parameters and display effects, including: Edge detection is performed on the original PWM drive signal to obtain a rising-edge time sequence and a falling-edge time sequence, and the duty cycle data of the drive signal is calculated according to the rising-edge time sequence and the falling-edge time sequence; Based on the duty cycle data, the original PWM drive signal is segmented to obtain a pre-charge segment waveform, a main drive segment waveform, and a stable segment waveform, and the pre-charge segment waveform, the main drive segment waveform, and the stable segment waveform are synchronized in time sequence; Amplitude modulation is performed on the pre-charge segment waveform to generate a positive compensation pulse in the line scan signal, and an inverted negative compensation pulse is generated according to the waveform parameters of the positive compensation pulse; Based on the positive compensation pulse and the negative compensation pulse, a compensation signal transfer function is established, and frequency-domain analysis is performed on the compensation signal transfer function through Laplace transform to obtain a compensation signal spectrum; According to the compensation signal spectrum, a crosstalk suppression filter is constructed, and the positive compensation pulse and the negative compensation pulse are filtered to obtain a corrected compensation pulse pair, and the corrected compensation pulse pair is respectively superimposed on the pre-charge segments of the line scan signal and the column scan signal to generate a complementary drive waveform pair; An orthogonal test is performed on the complementary drive waveform pair within a preset voltage range and a preset time range to establish a characteristic correlation matrix between drive parameters and display effects.

4. The method for optimizing the driving waveform of the small-pitch LED array according to claim 3, wherein The orthogonal test is performed on the complementary drive waveform pair within a preset voltage range and a preset time range to establish a characteristic correlation matrix between drive parameters and display effects, including: Orthogonal test factors are constructed for the compensation pulse amplitude and pulse width of the complementary drive waveform pair. The compensation pulse amplitude is divided into N test points within the target voltage range, and the pulse width is divided into M test points within the target time range to obtain N×M groups of test parameters; An orthogonal test table is generated based on the N×M groups of test parameters, and the complementary drive waveform pair is parameter-set according to the orthogonal test table to obtain N×M groups of experimental waveforms; Waveform analysis is performed on the current rising-edge change rate, overshoot amplitude, and crosstalk current of the N×M groups of experimental waveforms to obtain the dynamic characteristic parameters of each group of waveforms; Based on the dynamic characteristic parameters and the display quality evaluation function, a multi-dimensional characteristic space is constructed, and principal vector decomposition operation is performed on the multi-dimensional characteristic space to obtain a group of eigenvectors; According to the group of eigenvectors, a mapping relationship between drive parameters and display effects is established, least-squares fitting calculation is performed on the mapping relationship to obtain a correlation coefficient matrix, and singular value decomposition operation is performed on the correlation coefficient matrix to generate a characteristic correlation matrix between drive parameters and display effects.

5. The method for optimizing the driving waveform of the small-pitch LED array according to claim 1, wherein Calculating an optimal drive parameter combination according to the characteristic correlation matrix and the display quality evaluation function, and generating an LED array drive signal according to the optimal drive parameter combination, including: Construct a target optimization function based on the characteristic correlation matrix, set constraint conditions for the target optimization function, and use the compensation pulse amplitude range and pulse width range as the particle position boundaries to obtain the target search space; Initialize particles for the target search space, construct a chaotic sequence to perturb the initial particle swarm, obtain the initial solution space, and calculate the fitness value of each initial particle based on the display quality evaluation function; Calculate the adaptive inertia weight for the particles in the initial solution space, and dynamically adjust the inertia weight according to the fitness value to obtain the motion parameters of the particle swarm; Perform a velocity update operation on the particle swarm based on the motion parameters, and introduce a local search operator to update the particle positions to obtain the candidate solution set after iteration; Conduct a constraint check on the candidate solution set, process the out-of-bounds particles through the Lagrange multiplier method to obtain the effective solution set, and sort the effective solution set according to the display quality evaluation function; Extract the optimal particle position parameters from the effective solution set, decode and convert the optimal particle position parameters into compensation pulse parameters to generate the optimal drive parameter combination; Generate a clock signal by passing the optimal drive parameter combination through a digital phase-locked loop, and input the clock signal into a digital pulse width modulator to generate an LED array drive signal.

6. The method for optimizing the driving waveform of the small-pitch LED array according to claim 5, wherein The step of generating a clock signal by passing the optimal drive parameter combination through a digital phase-locked loop and inputting the clock signal into a digital pulse width modulator to generate an LED array drive signal includes: Perform digital encoding on the optimal drive parameter combination, convert the compensation pulse amplitude parameter into a frequency division coefficient, and convert the pulse width parameter into a phase control word to obtain the digital phase-locked loop control parameter; Construct a phase accumulator based on the digital phase-locked loop control parameter, perform a sine mapping operation on the accumulated output of the phase accumulator to obtain the reference signal of the phase frequency detector; Conduct a phase frequency comparison between the reference signal and the feedback signal, and filter the comparison result through a digital loop filter to obtain the control voltage of the numerically controlled oscillator; Input the control voltage into the numerically controlled oscillator for frequency modulation to generate a reference clock signal, and perform frequency division processing on the reference clock signal to obtain a feedback clock signal; Establish a counter for the digital pulse width modulation unit based on the reference clock signal, perform a comparison operation between the count value of the counter and the duty cycle data of the compensation pulse to obtain a pulse width modulation waveform, and perform level conversion and current amplification on the pulse width modulation waveform through a driver circuit to generate an LED array drive signal.

7. An apparatus for optimizing the driving waveform of a small-pitch LED array, characterized in that The device for executing the small-pitch LED array drive waveform optimization method according to any one of claims 1-6 includes: An acquisition module for collecting the spectral characteristics of the pixel points in the small-pitch LED array and constructing a mathematical model of the pixel point light emission characteristics; A construction module for performing parameter operations on the mathematical model of the pixel point light emission characteristics to obtain array display parameters, and constructing a display quality evaluation function according to the array display parameters; A test module for performing waveform compensation on the original PWM drive signal, generating a complementary drive waveform pair, and performing an orthogonal test based on the complementary drive waveform pair to establish a characteristic correlation matrix between drive parameters and display effects; A generation module for calculating an optimal drive parameter combination according to the characteristic correlation matrix and the display quality evaluation function, and generating an LED array drive signal according to the optimal drive parameter combination.

Images