A method for dimming an organic electroluminescent display panel

Abstract

The present invention provides a dimming method for an organic electroluminescent display panel, and belongs to the technical field of display panel dimming. The dimming method includes: data collection, data processing, model construction and optimization, model training and evaluation, and model application. Through the collected device parameters, the light intensity of the external environment, and the display content data, and the feature extraction work performed on this basis, the complex data is converted into a feature vector with representativeness and discrimination. Through the powerful learning ability and generalization ability of the multi-layer perceptron model, the complex nonlinear relationship between the device parameters and the external data and the dimming strategy is captured, and the particle swarm algorithm is introduced to optimize the hyperparameters. Based on the dimming parameters obtained by the optimized multi-layer perceptron model, the current, voltage, and pulse width are adjusted respectively, which solves the problems of intelligent local dimming and energy saving of the organic electroluminescent display panel, and realizes the intelligent local dimming and energy saving of the organic electroluminescent display panel.

Description

Technical Field

[0001] The present invention relates to the technical field of display panel dimming, and in particular to a method for dimming an organic electroluminescent display panel. Background Art

[0002] Organic electroluminescent display panel is a new type of display technology with the following characteristics: Organic electroluminescent display panel materials emit light when powered on, and do not require a backlight, so thinner and lighter display devices can be achieved. The brightness of each pixel can be controlled independently, thereby achieving a higher contrast ratio. At the same time, it can maintain good color performance over a wide viewing angle. With a response speed of milliseconds, smooth animation and video playback can be achieved. Organic electroluminescent display panels can completely turn off the corresponding pixels when displaying black, thereby saving a lot of electricity. The display is based on organic polymer materials and can be made into flexible and foldable screens. Organic electroluminescent display panel display technology is currently widely used in smart phones, tablets, TVs, wearable devices and other fields, and is constantly improving technology levels and reducing costs, and is expected to replace traditional LCD technology in the future.

[0003] Organic electroluminescent display panels still have some limitations in some aspects, including imperfect intelligent local dimming: the local dimming capability of organic electroluminescent display panels needs to be further improved, and the overall energy consumption is not low enough: the organic electroluminescent display panel can completely turn off the corresponding pixels when displaying black, thereby saving energy, but for most images and videos, the overall power consumption of the organic electroluminescent display panel is still higher than that of LCD display. Therefore, organic electroluminescent display panels still need to be improved in terms of intelligent local dimming and energy saving. Summary of the invention

[0004] The present invention provides an organic electroluminescent display panel dimming method, which is used to solve the defects of intelligent local dimming and energy saving in the prior art.

[0005] The present invention provides a method for dimming an organic electroluminescent display panel, the method comprising:

[0006] S1. Data collection: collect equipment parameters and external related data.

[0007] S2. Data processing: including missing value and outlier processing, data standardization, equipment parameter encoding and conversion, and external environment feature extraction.

[0008] S3. Model construction and optimization: Build a multi-layer perceptron model and optimize the hyperparameters of the multi-layer perceptron model through particle swarm algorithm.

[0009] S4. Model training and evaluation: Use the training set to train the multilayer perceptron model, and evaluate the performance of the multilayer perceptron model on the validation set.

[0010] S5. Intelligent local dimming is achieved through the trained multi-layer perceptron model, including:

[0011] S51: Obtain dimming parameters of different areas of the organic electroluminescent display panel through the trained multi-layer perceptron model.

[0012] S52: According to the dimming parameters output by the trained multi-layer perceptron model, the current, voltage and pulse width are adjusted respectively.

[0013] S53: Monitor brightness changes in real time, and continuously provide feedback and make adjustments to achieve local intelligent adjustment of the brightness of the organic electroluminescent display panel.

[0014] According to an organic electroluminescent display panel dimming method provided by the present invention, the device parameters include panel type, panel size, resolution, refresh rate and luminescent material characteristics. External related data include external light intensity and display content.

[0015] According to an organic electroluminescent display panel dimming method provided by the present invention, in step S2, the manner of processing missing values ​​and outliers includes: partially missing values ​​in device parameters are filled by the mean filling method. Partial missing values ​​in external related data are supplemented by linear interpolation. The outlier range is set by a box plot, outliers that confirm measurement errors are deleted, and true extreme values ​​are retained and marked.

[0016] According to an organic electroluminescent display panel dimming method provided by the present invention, in step S2, the formula for data standardization is:

[0017]

[0018] Among them, x is the original data, is the mean of the data, is the standard deviation of the data.

[0019] The present invention provides an organic electroluminescent display panel dimming method, characterized in that in step S2, the device parameter encoding and conversion method includes: using one-hot encoding to convert the panel type into a binary vector. The values ​​of the panel size and resolution are logarithmically transformed to reduce the magnitude difference of the data, and the formula is:

[0020]

[0021] Where x is the original panel size or resolution value and b is the base of the logarithm.

[0022] According to an organic electroluminescent display panel dimming method provided by the present invention, in step S2, the external environment feature extraction includes:

[0023] S21. Divide the external light intensity into multiple intervals and construct segmented features of the light intensity.

[0024] S22, extracting the average brightness of the displayed image: converting the image from the RGB color space to the HSV color space, summing the pixel values ​​of the brightness channel in the converted color space, and dividing the sum by the total number of pixels in the image to obtain the average brightness.

[0025] S23, extracting the contrast of the displayed image: calculating the grayscale image of the image, calculating the grayscale mean of the grayscale image, calculating the standard deviation of the grayscale image, and taking the ratio of the standard deviation to the grayscale mean as the contrast.

[0026] S24, extracting the average color saturation of the displayed image: summing the pixel values ​​of the saturation channel in HSV, and dividing the sum by the total number of pixels to obtain the average color saturation.

[0027] S25, extracting the brightness change rate between adjacent frames of the dynamic display content: the calculation formula of the brightness change rate between adjacent frames of the dynamic display content is:

[0028]

[0029] in, and are the average brightness of two adjacent frames of images.

[0030] S26, extracting the average color change amplitude between adjacent frames of the dynamic display content: For two adjacent frames of images, convert from RGB color space to Lab color space. For each pixel, calculate the Euclidean distance of its Lab color value in the two frames of images. Sum and average the Euclidean distances of all pixels to obtain the average color change amplitude.

[0031] According to an organic electroluminescent display panel dimming method provided by the present invention, in step S3, the construction of a multi-layer perceptron model includes:

[0032] S31. Construct a multi-layer perceptron model, including an input layer: receiving processed device parameters and external environment characteristics. Hidden layer: setting multiple hidden layers, each hidden layer including a preset number of neurons. Output layer: outputting current dimming, voltage dimming and pulse width dimming parameters.

[0033] S32. Select an activation function and use the Sigmoid function as the activation function of the multilayer perceptron model. The expression formula of the Sigmoid function is:

[0034]

[0035] In it, the input value x is between 0 and 1, and the output value f(x) varies smoothly between 0 and 1.

[0036] According to an organic electroluminescent display panel dimming method provided by the present invention, in step S3, the particle swarm algorithm optimizes the hyper parameters of the multi-layer perceptron model including:

[0037] a. Hyperparameter combinations are formed by learning rate, number of hidden layer neurons and regularization parameters.

[0038] b. Randomly generate a set of particles, and the position of each particle represents a set of hyperparameters.

[0039] c. Use the validation set data to evaluate the performance of the multilayer perceptron model under the hyperparameter combination represented by each particle as the fitness value.

[0040] d. Update the individual optimal position of each particle and the global optimal position of the entire particle swarm according to the fitness value. The update formula of the particle speed and position is:

[0041]

[0042]

[0043] in, is the velocity of particle i at the kth iteration, w is the inertia weight, and is the learning factor, and is a random number in the interval [0, 1]. is the individual optimal position of particle i up to the kth iteration, is the global optimal position of the entire particle swarm up to the kth iteration, is the position of particle i at the kth iteration.

[0044] According to an organic electroluminescent display panel dimming method provided by the present invention, in step S4, the preprocessed data is divided into a training set, a validation set and a test set, and the multi-layer perceptron model is trained using the training set data. The root mean square error is used as an evaluation index to evaluate the performance of the multi-layer perceptron model on the validation set, and the multi-layer perceptron model with the best performance is selected. The calculation formula of the root mean square error is:

[0045]

[0046] Among them, m is the number of samples in the validation set, is the true value, is the predicted value.

[0047] According to an organic electroluminescent display panel dimming method provided by the present invention, in step S52, the current dimming formula is:

[0048]

[0049] in, is the brightness produced by current dimming, is the current dimming coefficient, and I is the current value.

[0050] The voltage dimming formula is:

[0051]

[0052] in, is the brightness produced by voltage dimming, is the voltage dimming coefficient, V is the voltage value, and n is between 1 and 2.

[0053] The pulse width dimming formula is:

[0054]

[0055] in, is the average brightness achieved by pulse width dimming, is the maximum brightness and D is the duty cycle.

[0056] The present invention provides an organic electroluminescent display panel dimming method. By collecting device parameters, the light intensity of the external environment, and display content data, and performing feature extraction based on this, complex data can be converted into a representative and distinguishing feature vector. Through the powerful learning and generalization capabilities of the multi-layer perceptron model, the complex nonlinear relationship between device parameters, external data, and dimming strategies can be captured, and a particle swarm algorithm is introduced to optimize hyperparameters. The particle swarm algorithm simulates the foraging behavior of a flock of birds and quickly searches for the optimal solution in the hyperparameter space, thereby significantly improving the accuracy and stability of the multi-layer perceptron model. Based on the dimming parameters obtained by the optimized multi-layer perceptron model, the current, voltage, and pulse width are adjusted respectively, solving the problems of intelligent local dimming and energy saving of organic electroluminescent display panels, and realizing intelligent local dimming and energy saving of organic electroluminescent display panels. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0058] Figure 1 It is a schematic diagram of a flow chart of a method for dimming an organic electroluminescent display panel provided by an embodiment of the present invention;

[0059] Figure 2 It is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention.

[0060] Reference numerals:

[0061] 610, processor; 620, communication interface; 630, memory; 640, communication bus. DETAILED DESCRIPTION

[0062] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0063] Combine the following Figure 1-Figure 2 The present invention describes an organic electroluminescent display panel dimming method.

[0064] Figure 1 The present invention is a flowchart of an organic electroluminescent display panel dimming method provided by an embodiment of the present invention.

[0065] like Figure 1 As shown, an embodiment of the present invention provides a dimming method for an organic electroluminescent display panel, the method mainly comprising:

[0066] S1. Data collection: Collect device parameters and external related data. Device parameters include panel type, panel size, resolution, refresh rate and luminous material characteristics. External related data includes external light intensity and display content.

[0067] In this embodiment, the panel size includes length, width, thickness, etc. Resolution refers to the number of horizontal and vertical pixels. Refresh rate refers to the number of times the screen is refreshed per unit time. Luminous material properties include luminous efficiency and lifespan. External light intensity is obtained using a high-precision light sensor in lux. Display content includes average brightness, contrast, and color distribution of the image.

[0068] S2. Data processing: including missing value and outlier processing, data standardization, equipment parameter encoding and conversion, and external environment feature extraction.

[0069] The missing value and outlier processing includes: for a small number of missing values ​​in the equipment parameters, the mean filling method is used to fill them. For a small number of missing values ​​in external related data, the linear interpolation method is used to supplement them. The outlier value range is set through the box plot, and the outlier values ​​that are confirmed to be measurement errors are deleted, and the real extreme values ​​are retained and marked.

[0070] The equipment parameters and external condition data are standardized, and the standardized formula is:

[0071]

[0072] Among them, x is the original data, is the mean of the data, is the standard deviation of the data.

[0073] Device parameter encoding and conversion, including panel type encoding, size and resolution conversion, uses one-hot encoding to convert the panel type into a binary vector. Logarithmic transformation is performed on the values ​​of panel size and resolution to reduce the magnitude difference of the data. The formula is:

[0074]

[0075] Where x is the original panel size or resolution value and b is the base of the logarithm.

[0076] External environment feature extraction includes:

[0077] S21. Divide the external light intensity into multiple intervals and construct segmented features of the light intensity.

[0078] S22, extracting the average brightness of the displayed image: converting the image from the RGB color space to the HSV color space, summing the pixel values ​​of the brightness channel in the converted color space, and dividing the sum by the total number of pixels in the image to obtain the average brightness.

[0079] S23, extracting the contrast of the displayed image: calculating the grayscale image of the image, calculating the grayscale mean of the grayscale image, calculating the standard deviation of the grayscale image, and taking the ratio of the standard deviation to the mean as the contrast.

[0080] S24, extracting the average color saturation of the displayed image: summing the pixel values ​​of the saturation channel in HSV, and dividing the sum by the total number of pixels to obtain the average color saturation.

[0081] S25, extracting the brightness change rate between adjacent frames of the dynamic display content: the calculation formula of the brightness change rate between adjacent frames of the dynamic display content is:

[0082]

[0083] in, and are the average brightness of two adjacent frames of images.

[0084] S26, extracting the average color change amplitude between adjacent frames of the dynamic display content: For two adjacent frames of images, convert from RGB color space to Lab color space. For each pixel, calculate the Euclidean distance of its Lab color value in the two frames of images. Sum and average the Euclidean distances of all pixels to obtain the average color change amplitude.

[0085] In this embodiment, the collected device parameters, the light intensity of the external environment, and the display content data provide a rich information basis for the subsequent construction and optimization of the multi-layer perceptron model. The feature extraction work performed on this basis can transform complex data into representative and discriminative feature vectors, so that the multi-layer perceptron model can more effectively learn and understand the potential laws in the data.

[0086] S3. Model construction and optimization: Build a multi-layer perceptron model and optimize the hyperparameters of the multi-layer perceptron model through particle swarm algorithm.

[0087] The construction of the multi-layer perceptron model includes:

[0088] S31. Construct a multi-layer perceptron model, including an input layer: receiving processed device parameters and external environment characteristics. Hidden layer: setting multiple hidden layers, each with a preset number of neurons. Output layer: outputting current dimming, voltage dimming and pulse width dimming parameters.

[0089] S32. Select an activation function and use the Sigmoid function as the activation function of the multilayer perceptron model. The expression formula of the Sigmoid function is:

[0090]

[0091] In it, the input value x is between 0 and 1, and the output value f(x) varies smoothly between 0 and 1.

[0092] The particle swarm algorithm optimizes the hyperparameters of the multilayer perceptron model including:

[0093] a. Hyperparameter combinations are formed by learning rate, number of hidden layer neurons and regularization parameters.

[0094] b. Randomly generate a set of particles, and the position of each particle represents a set of hyperparameters.

[0095] c. Use the validation set data to evaluate the performance of the multilayer perceptron model under the hyperparameter combination represented by each particle as the fitness value.

[0096] d. Update the individual optimal position of each particle and the global optimal position of the entire particle swarm according to the fitness value. The update formula of the particle speed and position is:

[0097]

[0098]

[0099] in, is the velocity of particle i at the kth iteration, w is the inertia weight, and is the learning factor, and is a random number in the interval [0, 1]. is the individual optimal position of particle i up to the kth iteration, is the global optimal position of the entire particle swarm up to the kth iteration, is the position of particle i at the kth iteration.

[0100] In this embodiment, the constructed multi-layer perceptron model has strong learning and generalization capabilities, and can capture the complex nonlinear relationship between device parameters, external data and dimming parameters. At the same time, the particle swarm algorithm is introduced to optimize the hyperparameters. The particle swarm algorithm simulates the foraging behavior of bird flocks and quickly searches for the optimal solution in the hyperparameter space, thereby significantly improving the accuracy and stability of the multi-layer perceptron model.

[0101] S4. Model training and evaluation: Use the training set to train the multilayer perceptron model, and evaluate the performance of the multilayer perceptron model on the validation set.

[0102] The preprocessed data is divided into training set, validation set and test set, and the multilayer perceptron model is trained using the training set data. The root mean square error is used as the evaluation index to evaluate the performance of the multilayer perceptron model on the validation set, and the multilayer perceptron model with the best performance is selected. The calculation formula of the root mean square error is:

[0103]

[0104] Among them, m is the number of samples in the validation set, is the true value, is the predicted value.

[0105] S5. Intelligent local dimming is achieved through the trained multi-layer perceptron model, including:

[0106] S51: Obtain dimming parameters of different areas of the organic electroluminescent display panel through the trained multi-layer perceptron model.

[0107] S52: According to the dimming parameters output by the multi-layer perceptron model, the current, voltage and pulse width are adjusted respectively.

[0108] S53: Monitor brightness changes in real time, continuously provide feedback and make adjustments, and realize local intelligent adjustment of the brightness of the organic electroluminescent display panel.

[0109] In this embodiment, the processed data is received through the input layer of the multi-layer perceptron model, the multi-layer perceptron structure is constructed through the hidden layer, and the dimming strategy is learned using the back propagation algorithm. The dimming parameters, including current, voltage and pulse width parameters, are output through the output layer, thereby realizing local intelligent adjustment of the brightness of the organic electroluminescent display panel.

[0110] Among them, the formula for current dimming is:

[0111]

[0112] in, is the brightness produced by current dimming, is the current dimming coefficient, and I is the current value.

[0113] The voltage dimming formula is:

[0114]

[0115] in, is the brightness produced by voltage dimming, is the voltage dimming coefficient, V is the voltage value, and n is between 1 and 2.

[0116] The pulse width dimming formula is:

[0117]

[0118] in, is the average brightness achieved by pulse width dimming, is the maximum brightness and D is the duty cycle.

[0119] Back propagation training of the multilayer perceptron model includes:

[0120] Forward propagation: propagate the input data forward from the input layer to the output layer layer by layer, and calculate the neuron activation value of each layer.

[0121] Error calculation: Calculate the error between the output value of the output layer neuron and the expected output value, and calculate the local error gradient of each neuron.

[0122] Back propagation: propagate the error back from the output layer to the input layer layer by layer, and update each connection weight and neuron bias according to the gradient descent method.

[0123] Iterative optimization: Repeat forward propagation and back propagation until the multilayer perceptron model error converges to a small enough size.

[0124] In this embodiment, the dimming parameters obtained based on the optimized multi-layer perceptron model can realize accurate local dimming of the organic electroluminescent display panel. In terms of current dimming, by accurately adjusting the current flowing through the panel, the brightness is finely controlled, the energy consumption increase and component aging caused by excessive current are avoided, and the display effect under low current is also guaranteed. Voltage dimming effectively reduces the power consumption of the panel while ensuring the display quality. According to different scenarios and requirements, the voltage applied to the panel is intelligently adjusted so that the brightness can be perfectly matched with the environment and the display content. Pulse width dimming further enhances the flexibility and accuracy of dimming. By accurately controlling the duty cycle of the pulse, the brightness can be adjusted quickly and accurately, especially in dynamic display scenes, and can effectively reduce the flicker and afterimage phenomenon of the screen. By combining these three dimming methods, we have successfully realized the local dimming of the organic electroluminescent display panel, bringing many significant effects. In terms of display quality, the contrast has been significantly improved, the colors are more vivid and accurate, and the image details are clearer, bringing users a more realistic and comfortable visual experience. At the same time, breakthrough progress has been made in energy consumption management, which has greatly reduced the overall power consumption of the panel and extended the battery life of the equipment, which is in line with the current development trend of energy conservation and environmental protection.

[0125] Figure 2 It is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention.

[0126] like Figure 2 As shown, the electronic device may include: a processor 610, a communication interface 620, a memory 630 and a communication bus 640, wherein the processor 610, the communication interface 620 and the memory 630 communicate with each other through the communication bus 640. The processor 610 may call the logic instructions in the memory 630 to execute a dimming method for an organic electroluminescent display panel.

[0127] In addition, the logic instructions in the above-mentioned memory 630 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0128] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0129] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0130] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for dimming an organic electroluminescent display panel, characterized in that: The method includes: S1. Data collection: collect equipment parameters and external related data; S2. Data processing: including missing value and outlier processing, data standardization, equipment parameter encoding and conversion, and external environment feature extraction; S3, model construction and optimization: construct a multi-layer perceptron model, and optimize the hyperparameters of the multi-layer perceptron model by particle swarm algorithm; The construction of the multi-layer perceptron model includes: S31. Construct a multi-layer perceptron model, including an input layer: receiving processed device parameters and external environment characteristics; a hidden layer: setting multiple hidden layers, each hidden layer including a preset number of neurons; an output layer: outputting current dimming, voltage dimming and pulse width dimming parameters; S32, select an activation function, and use the Sigmoid function as the activation function of the multilayer perceptron model. The expression formula of the Sigmoid function is: Among them, the input value x is between 0 and 1, and the output value f(x) changes smoothly between 0 and 1; The particle swarm algorithm optimizes the hyper parameters of the multilayer perceptron model including: a. Hyperparameter combinations are formed by learning rate, number of hidden layer neurons and regularization parameters; b. Randomly generate a set of particles, where the position of each particle represents a set of hyperparameters; c. Use the validation set data to evaluate the performance of the multilayer perceptron model under the hyperparameter combination represented by each particle as the fitness value; d. Update the individual optimal position of each particle and the global optimal position of the entire particle swarm according to the fitness value. The update formula of the particle speed and position is: in, is the velocity of particle i at the kth iteration, w is the inertia weight, and is the learning factor, and is a random number in the interval [0, 1]. is the individual optimal position of particle i up to the kth iteration, is the global optimal position of the entire particle swarm up to the kth iteration, is the position of particle i at the kth iteration; S4, model training and evaluation: using the training set to train the multilayer perceptron model, and evaluating the performance of the multilayer perceptron model on the validation set; S5. Intelligent local dimming is achieved through the trained multi-layer perceptron model, including: S51: Obtaining dimming parameters of different areas of the organic electroluminescent display panel through the trained multi-layer perceptron model; S52: adjusting the current, voltage and pulse width respectively according to the dimming parameters output by the trained multi-layer perceptron model; S53: Real-time monitoring of brightness changes, continuous feedback adjustment, and implementation of local intelligent adjustment of the brightness of the organic electroluminescent display panel.

2. The method for dimming an organic electroluminescent display panel according to claim 1, characterized in that: The device parameters include panel type, panel size, resolution, refresh rate and luminescent material characteristics; the external related data include external light intensity and display content.

3. The method for dimming an organic electroluminescent display panel according to claim 1, characterized in that: In step S2, the missing values ​​and outlier processing methods include: filling some missing values ​​in the device parameters with the mean filling method; supplementing some missing values ​​in the external related data with the linear interpolation method; setting the outlier value range through the box plot, deleting outliers that confirm measurement errors, and retaining and marking the real extreme values.

4. The method for dimming an organic electroluminescent display panel according to claim 1, characterized in that: In step S2, the formula for data standardization is: in, is the original data, is the mean of the data, is the standard deviation of the data, is the standardized data.

5. The method for dimming an organic electroluminescent display panel according to claim 1, characterized in that: In step S2, the device parameter encoding and conversion method includes: using one-hot encoding to convert the panel type into a binary vector; performing a logarithmic transformation on the values ​​of the panel size and resolution to reduce the magnitude difference of the data, and the formula is: Where z is the original panel size or resolution and b is the base of the logarithm.

6. The method for dimming an organic electroluminescent display panel according to claim 1, characterized in that: In step S2, the external environment feature extraction includes: S21, dividing the external light intensity into multiple intervals, and constructing segmented features of the light intensity; S22, extracting the average brightness of the displayed image: converting the image from the RGB color space to the HSV color space; summing the pixel values ​​of the brightness channel in the converted color space; dividing the sum by the total number of pixels of the image to obtain the average brightness; S23, extracting the contrast of the displayed image: calculating a grayscale image of the image; calculating the grayscale mean of the grayscale image; calculating the standard deviation of the grayscale image; and taking the ratio of the standard deviation to the grayscale mean as the contrast; S24, extracting the average color saturation of the displayed image: summing the pixel values ​​of the saturation channel in the HSV; dividing the obtained sum by the total number of pixels to obtain the average color saturation; S25, extracting the brightness change rate between adjacent frames of the dynamic display content: the calculation formula of the brightness change rate between adjacent frames of the dynamic display content is: in, and are the average brightness of two adjacent frames of images; S26. Extract the average color change amplitude between adjacent frames of the dynamic display content: for two adjacent frames of images, convert from RGB color space to Lab color space; for each pixel, calculate the Euclidean distance of its Lab color value in the two frames of images; sum and average the Euclidean distances of all pixels to obtain the average color change amplitude.

7. The method for dimming an organic electroluminescent display panel according to claim 1, characterized in that: In step S4, the preprocessed data is divided into a training set, a validation set and a test set, and the multilayer perceptron model is trained using the training set data; the root mean square error is used as an evaluation index to evaluate the performance of the multilayer perceptron model on the validation set, and the multilayer perceptron model with the best performance is selected; wherein the calculation formula of the root mean square error is: Where m is the number of samples in the validation set, is the true value, is the predicted value.

8. The method for dimming an organic electroluminescent display panel according to claim 1, characterized in that: In step S52, the current dimming formula is: in, is the brightness produced by current dimming, is the current dimming coefficient, I is the current value; The voltage dimming formula is: in, is the brightness produced by voltage dimming, is the voltage dimming coefficient, V is the voltage value, and n is between 1 and 2; The pulse width dimming formula is: in, is the average brightness achieved by pulse width dimming, is the maximum brightness and D is the duty cycle.

Images