LED light-emitting assembly driving IC control method and system and driving IC

Abstract

The invention belongs to the technical field of display, and discloses an LED light-emitting assembly driving IC control method and system and a driving IC, and the method comprises the steps: receiving an external control signal; preprocessing the received external control signal to obtain a standardized digital control signal; collecting working state data of the LED light-emitting assembly; preprocessing the working state data of the LED light-emitting assembly to obtain a working state feature data set; training based on the standardized digital control signal and the working state characteristic data set to obtain a strategy optimization model, and predicting to obtain a driving IC adjusting signal; the working state of the LED light-emitting assembly is adjusted according to the driving IC adjusting signal; real-time light intensity output of the LED light-emitting assembly is monitored through the light sensor, the real-time light intensity output is compared with preset light intensity output, and whether deviation is generated or not is judged; the service life of the LED light-emitting assembly is prolonged, and the performance of the whole system is improved.

Description

Technical Field

[0001] The present invention relates to the field of display technology, and more specifically, to a method and system for controlling a LED light emitting component driver IC, and a driver IC. Background Art

[0002] The patent with application publication number CN105405412A discloses a backlight drive control method and system, which is applied to a display terminal with a liquid crystal display screen. The display terminal includes an MCU and a backlight source. The backlight source includes multiple light bars formed by LED lights and multiple driver ICs. Each driver IC controls one or more light bars. The method includes the following steps: when the MCU receives the electrical signal of each LED light, demodulate the electrical signal of each LED light to obtain the brightness value of each LED light; compare the brightness values ​​of each LED light in each driver IC to obtain the maximum brightness of the LED light in each driver IC; divide the maximum brightness of the LED light in each driver IC into several brightness intervals, determine the current value corresponding to each brightness interval, and control the brightness of the LED light in the driver IC according to the current value. The present invention improves the energy saving effect of LocalDimming drive and the contrast of screen display.

[0003] The existing LED light-emitting component driver IC control method, system and driver IC have the following main problems: The working status data of the LED light-emitting component is not divided into multiple signal channels according to the data type. Different types of data may interfere with each other, making data analysis difficult and unable to accurately reflect the actual working status of the LED light-emitting component. The lack of wavelet decomposition will make it impossible to accurately extract the main trends and rapid changes or detailed information of the data, which limits the in-depth understanding of the data characteristics and may lead to inaccurate analysis of the performance of the LED light-emitting component. The covariance matrix and eigenvalue decomposition are not used for multi-channel joint analysis, which will not be able to capture the cross-channel correlation between signal channels, and thus multi-channel joint denoising cannot be achieved, which may lead to the loss of important information or the retention of too much noise during the processing process. The soft thresholding method is not considered for thresholding the eigenvalue matrix, which will not be able to effectively remove noise features while retaining important features. The low-frequency approximate coefficient matrix and the high-frequency detail coefficient matrix after denoising are not combined and reconstructed through the inverse wavelet transform, which will not be able to restore a clearer and more accurate working status feature data set. Failure to utilize the bacterial foraging optimization algorithm may result in inefficiency in exploring and optimizing the configuration of the driver IC regulation signal. Traditional methods may require longer computing time and higher computing resources to find a near-optimal solution; lacking an adjustment mechanism for the adaptive step size formula, the algorithm may fall into a local optimal solution too early and be unable to continue exploring a more optimal solution globally, resulting in a limited degree of optimization of the driver IC regulation signal configuration and the inability of the LED light-emitting component to achieve the optimal light intensity output; without a mechanism to gradually reduce the step size, the algorithm may not be able to search the solution space finely, resulting in insufficient search accuracy, affecting the precise control of the LED light-emitting component light intensity output, and reducing system stability and reliability; without the support of the regulation factor restriction formula and the random walk formula, the algorithm may not be able to effectively balance the needs of the two stages of exploration and development, resulting in an inability to quickly traverse the solution space in the early stages of the search, and an inability to finely search for potential optimal solutions in the later stages of the search.

[0004] In view of this, the present invention proposes a LED light emitting component driver IC control method, system and driver IC to solve the above problems. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art and to achieve the above-mentioned purpose, the present invention provides the following technical solution: a method for controlling an LED light-emitting component driver IC, comprising: S1, receiving external control signal; S2, preprocessing the received external control signal to obtain a standardized digital control signal; S3, collecting the working state data of the LED light emitting component; preprocessing the working state data of the LED light emitting component to obtain a working state feature data set; S4. Based on the standardized digital control signal and the working state characteristic data set, the acquisition strategy optimization model is trained to predict the driver IC adjustment signal; and the working state of the LED light-emitting component is adjusted according to the driver IC adjustment signal; S5. Monitor the real-time light intensity output of the LED light-emitting component through a light sensor, compare the real-time light intensity output with the preset light intensity output, and determine whether a deviation occurs; if a deviation occurs, calculate the light intensity deviation amount, and further optimize the driver IC adjustment signal based on the light intensity deviation amount until the light intensity output of the LED light-emitting component reaches the optimal level.

[0006] Furthermore, the external control signal includes a brightness control signal, a color temperature control signal, a working mode control signal, a current control signal, a voltage control signal, an environment monitoring signal and a safety control signal; The working status data of the LED light-emitting component includes electrical parameter data, temperature parameter data, optical parameter data and operating time; the electrical parameter data includes current and voltage; the temperature parameter data includes the LED light-emitting component temperature and the driver IC temperature; the optical parameter data includes brightness and color temperature.

[0007] Furthermore, the method of preprocessing the received external control signal to obtain a standardized digital control signal includes: The received external control signal is converted into a digital signal using an analog-to-digital converter, and is normalized to a uniform digital range by performing standardization processing; the normalized digital signal is denoised using a low-pass filter to obtain a denoised digital signal; the denoised digital signal is decoded into an operating mode instruction, and the obtained operating mode instruction is mapped into control parameters to obtain a standardized digital control signal.

[0008] Furthermore, the method of preprocessing the working state data of the LED light-emitting component to obtain a working state feature data set includes: The working state data of the LED light-emitting component is divided into N signal channels according to different data types, and each signal channel is decomposed by wavelet to obtain low-frequency approximate coefficients and high-frequency detail coefficients; the working state data of the LED light-emitting component of the N signal channels are jointly analyzed through covariance matrix and eigenvalue decomposition, the cross-channel correlation between the N signal channels is captured, and multi-channel joint denoising is performed; the denoised low-frequency approximate coefficient matrix and high-frequency detail coefficient matrix are combined and reconstructed through inverse wavelet transform, and then the working state feature data set is obtained; the specific method steps are as follows: S41, dividing the working state data of the LED light-emitting component into N signal channels according to different data types, each signal channel containing a group of independent time series signals; performing timestamp alignment and normalization processing on the independent time series signals contained in the N signal channels, and integrating them into a time series vector containing the N signal channels; for the independent time series signals contained in each signal channel, using discrete wavelet transform to perform wavelet decomposition, and decompose them layer by layer into low-frequency approximate coefficients and high-frequency detail coefficients; S42, collect the high-frequency detail coefficients obtained by wavelet decomposition of N signal channels at each layer, and construct a high-frequency detail transformation matrix for joint analysis between channels; the high-frequency detail transformation matrix is ;in, For the first signal channel The high-frequency detail coefficient matrix obtained by wavelet decomposition of the layer; For the The signal channel is The high-frequency detail coefficient matrix obtained by wavelet decomposition of the layer; For the The signal channel is The high-frequency detail coefficient matrix obtained by wavelet decomposition of the layer; S43, calculate the covariance matrix of each layer of high-frequency detail coefficient matrix ;in, For the The number of high-frequency detail coefficients in the layer high-frequency detail coefficient matrix; is the high frequency detail coefficient matrix The transposed matrix of is the transposition operator; the covariance matrix of each layer of high-frequency detail coefficient matrix obtained Perform eigenvalue decomposition; the covariance matrix after eigenvalue decomposition is ;in, is the covariance matrix after eigenvalue decomposition; is the covariance matrix The eigenvector matrix of is the covariance matrix The eigenvalue matrix of is the covariance matrix The transposed matrix of the eigenvector matrix of ; S44, using soft threshold processing method, the eigenvalue matrix Perform thresholding; preset threshold , the eigenvalue matrix The feature value is less than the preset threshold Set the part of to zero and get the new eigenvalue matrix ; The threshold processing formula is: ;in, After thresholding eigenvalues; is a symbolic function, representing the eigenvalue The positive and negative ; To keep The characteristic value of The eigenvalues ​​of are set to zero; is the preset threshold, which is obtained through the threshold adjustment formula; the threshold adjustment formula is ;in is the standard deviation of the noise; log is the sign of the logarithmic function; S45, through the new eigenvalue matrix , get the denoised covariance matrix ; Based on the denoised covariance matrix , use the reconstruction formula to calculate the high-frequency detail coefficient matrix after denoising; the reconstruction formula is ;in, For Take the square root of each term; S46, collecting low-frequency approximate coefficients obtained by wavelet decomposition of N signal channels at each layer, constructing a low-frequency approximate transformation matrix for joint analysis between channels, repeating steps S42-S45, and then obtaining a denoised low-frequency approximate coefficient matrix; S47. Combining and reconstructing the denoised low-frequency approximate coefficient matrix and the high-frequency detail coefficient matrix through inverse wavelet transform, thereby obtaining a working status feature data set.

[0009] Furthermore, the training method of the strategy optimization model includes: The data set is divided into a training set, a validation set and a test set; the data set includes a historical standardized digital control signal and a working state feature data set and a corresponding driver IC adjustment signal; a strategy optimization model is constructed, and the strategy optimization model includes an input layer, a hidden layer and an output layer; the hidden layer uses a ReLU activation function; the input layer of the model is used to input the historical standardized digital control signal and the working state feature data set; the output layer of the model is used to output the driver IC adjustment signal, and the predicted value is output through a neuron; the strategy optimization model is a fully connected neural network model; Use mean absolute error as the loss function to measure the error between the model's predicted value and the actual value; Use the training set data to train the model and minimize the loss function using the Adam optimizer. Use the validation set to evaluate the performance of the model and tune the model's hyperparameters until the model performance stops improving or reaches the preset stopping condition. The test set is used to evaluate the performance of the model in the prediction task. The current standardized digital control signal and working state feature data set are input into the trained strategy optimization model to obtain the driver IC adjustment signal.

[0010] Furthermore, the method for adjusting the working state of the LED light emitting component according to the driver IC adjustment signal includes: The driver IC predicts the driver IC regulation signal through the strategy optimization model, and the driver IC regulation signal includes the current, voltage and PWM signal regulation parameters required by the LED light-emitting component; the driver IC regulation signal is converted into an actual output signal, and the working state of the LED light-emitting component is adjusted according to the actual output signal.

[0011] Furthermore, the method of comparing the real-time light intensity output with the preset light intensity output to determine whether a deviation occurs includes: If the real-time light intensity output is greater than or less than the preset light intensity output, it is determined that a deviation occurs; If the real-time light intensity output is equal to the preset light intensity output, it is determined that no deviation occurs.

[0012] Furthermore, if a deviation occurs, the light intensity deviation is calculated, and the driver IC adjustment signal is further optimized based on the light intensity deviation until the light intensity output of the LED light-emitting component reaches the optimal value. The method includes: The driving IC adjustment signal is further optimized based on the light intensity deviation through the bacterial foraging optimization algorithm until the light intensity output of the LED light-emitting component reaches the optimal value; the specific method steps are as follows: S81. Calculate the light intensity deviation between the real-time light intensity output of the LED light-emitting component and the preset light intensity output using a light intensity deviation calculation formula; the light intensity deviation calculation formula is: ;in, The preset light intensity output; It is real-time light intensity output; S82, further optimize the driver IC adjustment signal based on the light intensity deviation through the bacterial foraging optimization algorithm until the light intensity output of the LED light-emitting component reaches the optimal value; initialize the bacterial position, each bacterium represents a driver IC adjustment signal configuration; preset the population size to , each bacterial position is represented by a vector, denoted as ;in, For the Positional signals of bacteria; is the index of bacteria; the location of each bacterium Corresponding to a driver IC adjustment signal configuration, the real-time light intensity output of the LED is calculated through this configuration , and further calculate the light intensity deviation : ; S83. Define a fitness function to evaluate the quality of each bacterium to minimize the light intensity deviation; the fitness function is: ;in, For the The fitness value of each bacterium; S84, bacteria explore the solution space by swimming, and update their positions according to the current fitness value and the state of neighboring bacteria through the swimming update formula to find a better solution; the swimming update formula is: ;in, For the new locations of bacteria; For the the old location of the bacteria; is the step size factor that controls the swimming amplitude; is the global optimal position, that is, the driver IC adjustment signal configuration corresponding to the current optimal solution; For a between The random factor between S85. Dynamically adjust the step length factor for controlling the swimming amplitude through an adaptive step length formula; the adaptive step length formula is: ;in, is the dynamically adjusted step size factor; is the light intensity deviation of the current optimal solution; is the maximum light intensity deviation; A regulating factor for controlling the step size adjustment amplitude; S86. Limit the adjustment factor of the control step adjustment range by using the adjustment factor limiting formula; the adjustment factor limiting formula is: ;in, is the adjustment factor after restriction; is the number of current bacterial positions; The number of times the current bacterial position is updated; is the constant affecting the number of bacterial positions; Update influence constants for bacterial positions; S87. Use the random walk formula to simulate the random walk of bacteria in an unknown area; the random walk formula is: ;in, is the random walk step factor, which is used to control the jumping amplitude of bacteria in space; It is the optimal position for the bacterial colony; S88. After updating the position of each bacterium, recalculate its fitness, that is, calculate the new light intensity deviation, and evaluate its new fitness value; update the current optimal solution by comparing the fitness values; if the fitness of a certain bacterium is better, that is, the light intensity deviation is smaller, then update the position and fitness of the optimal solution; preset the minimum threshold of the light intensity deviation, when the light intensity deviation is The algorithm is stopped when the minimum threshold of the light intensity deviation is reached; the final optimized driver IC adjustment signal configuration is output, that is, the adjustment signal corresponding to the minimum light intensity deviation, at which time the light intensity output of the LED light-emitting component reaches the optimal level.

[0013] An LED light emitting component driving IC control system, comprising: A signal receiving module, used for receiving an external control signal; A signal processing module, used for preprocessing the received external control signal to obtain a standardized digital control signal; The data acquisition and processing module is used to collect the working state data of the LED light-emitting component; pre-process the working state data of the LED light-emitting component to obtain a working state characteristic data set; The optimization strategy module trains and acquires the strategy optimization model based on the standardized digital control signal and the working state characteristic data set, and predicts the driver IC adjustment signal; according to the driver IC adjustment signal, the working state of the LED light-emitting component is adjusted; The feedback adjustment module monitors the real-time light intensity output of the LED light-emitting component through a light sensor, compares the real-time light intensity output with the preset light intensity output, and determines whether a deviation occurs; if a deviation occurs, the light intensity deviation amount is calculated, and the driver IC adjustment signal is further optimized based on the light intensity deviation amount until the light intensity output of the LED light-emitting component reaches the optimal level.

[0014] A driver IC is located on an LED light-emitting component and has a built-in circuit for implementing a method for controlling a driver IC of an LED light-emitting component. The driver IC can receive a driver IC adjustment signal from a control system of a driver IC of an LED light-emitting component, convert the driver IC adjustment signal into an actual output signal, and adjust the working state of the LED light-emitting component according to the actual output signal; when the actual output signal is executed, a method for controlling a driver IC of an LED light-emitting component is implemented.

[0015] The present invention provides a method, system and driver IC control system for an LED light emitting component, which have the following beneficial effects: The working status data of the LED light-emitting component is divided into multiple signal channels according to the data type, so that different types of data can be processed independently, avoiding interference between data types; through wavelet decomposition, the main trend (low-frequency approximate coefficient) and fast-changing or detailed information (high-frequency detail coefficient) of the working status data of the LED light-emitting component can be accurately extracted, which is helpful for a deeper understanding of the data characteristics; multi-channel joint analysis is performed using covariance matrix and eigenvalue decomposition to capture the cross-channel correlation between signal channels, thereby realizing multi-channel joint denoising; compared with single-channel denoising, it is more effective and can more accurately retain useful information and remove noise; the soft threshold processing method is used to threshold the eigenvalue matrix, which can effectively remove noise features while retaining important features, thereby improving the signal-to-noise ratio of the data; the denoised low-frequency approximate coefficient matrix and high-frequency detail coefficient matrix are combined and reconstructed by inverse wavelet transform, which can restore a clearer and more accurate working status feature data set; By using the bacterial foraging optimization algorithm and simulating the swimming and foraging behavior of bacteria in the solution space, the configuration of the driver IC regulation signal can be explored and optimized quickly and efficiently; an adaptive step size formula is introduced, which can dynamically adjust the size of the step size factor according to the light intensity deviation of the current optimal solution, the maximum light intensity deviation and the regulation factor that controls the step size adjustment amplitude; it is helpful to quickly traverse the solution space in the early stage of the search to avoid falling into the local optimal solution too early; and in the later stage of the search, the step size can be gradually reduced to improve the accuracy of the local search; through the regulation factor restriction formula and the random walk formula, the method can balance the needs of the two stages of exploration and development; by continuously optimizing the configuration of the driver IC regulation signal, the light intensity output of the LED light-emitting component can be accurately controlled to reach the preset optimal value; by optimizing the configuration of the driver IC regulation signal, the light intensity deviation of the LED light-emitting component can be reduced, and the stability and reliability of the system can be improved; it is helpful to extend the service life of the LED light-emitting component, reduce the maintenance cost of the system, and improve the performance of the overall system. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A schematic diagram of a LED light emitting component driver IC control method, system and driver IC electronic method flow chart of the present invention; Figure 2 The present invention is a schematic diagram of the structure of an LED light emitting component driver IC control method, system and driver IC storage system. DETAILED DESCRIPTION

[0017] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only 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.

[0018] Example 1 See also Figure 1 In this embodiment, a method for controlling an LED light emitting component driving IC includes: S1, receiving external control signal; S2, preprocessing the received external control signal to obtain a standardized digital control signal; S3, collecting the working state data of the LED light emitting component; preprocessing the working state data of the LED light emitting component to obtain a working state feature data set; S4. Based on the standardized digital control signal and the working state characteristic data set, the acquisition strategy optimization model is trained to predict the driver IC adjustment signal; and the working state of the LED light-emitting component is adjusted according to the driver IC adjustment signal; S5. Monitor the real-time light intensity output of the LED light-emitting component through a light sensor, compare the real-time light intensity output with the preset light intensity output, and determine whether a deviation occurs; if a deviation occurs, calculate the light intensity deviation amount, and further optimize the driver IC adjustment signal based on the light intensity deviation amount until the light intensity output of the LED light-emitting component reaches the optimal level.

[0019] The external control signals include brightness control signal, color temperature control signal, working mode control signal, current control signal, voltage control signal, environment monitoring signal and safety control signal; The working status data of the LED light-emitting component includes electrical parameter data, temperature parameter data, optical parameter data and operating time; the electrical parameter data includes current and voltage; the temperature parameter data includes the LED light-emitting component temperature and the driver IC temperature; the optical parameter data includes brightness and color temperature.

[0020] The working status data of the LED light-emitting component is collected through embedded sensors, including current sensors, voltage sensors, temperature sensors and light sensors.

[0021] The method of preprocessing the received external control signal to obtain a standardized digital control signal includes: An analog-to-digital converter is used to convert the received external control signal into a digital signal, and the signal is normalized to a uniform digital range by performing standardization processing. A low-pass filter is used to perform denoising processing on the normalized digital signal to obtain a denoised digital signal. The denoised digital signal is decoded into an operating mode instruction, and the obtained operating mode instruction is mapped into control parameters, such as brightness level, color temperature setting, current size, etc., to obtain a standardized digital control signal.

[0022] The method of preprocessing the working state data of the LED light emitting component to obtain the working state feature data set includes: The working state data of the LED light-emitting component is divided into N signal channels according to different data types, and each signal channel is decomposed by wavelet to obtain low-frequency approximate coefficients and high-frequency detail coefficients; the working state data of the LED light-emitting component of the N signal channels are jointly analyzed through covariance matrix and eigenvalue decomposition, the cross-channel correlation between the N signal channels is captured, and multi-channel joint denoising is performed; the denoised low-frequency approximate coefficient matrix and high-frequency detail coefficient matrix are combined and reconstructed through inverse wavelet transform, and then the working state feature data set is obtained; the specific method steps are as follows: S41, dividing the working status data of the LED light-emitting component into N signal channels according to different data types, each signal channel contains a group of independent time series signals; performing timestamp alignment and normalization processing on the independent time series signals contained in the N signal channels, and integrating them into a time series vector containing N signal channels; the time series vector is ;in, For the signal channel The working status data of the LED light-emitting components at the moment; For the In the channel The working status data of the LED light-emitting components at the moment; and is the index of the signal channel type, ; is the index of the moment; for the independent time series signal contained in each signal channel, discrete wavelet transform is used to perform wavelet decomposition, and it is decomposed layer by layer into low-frequency approximate coefficients and high-frequency detail coefficients; The low-frequency approximation coefficient reflects the main trends and large-scale characteristics of the working status data of the LED light-emitting component; the high-frequency detail coefficient reflects the rapid changes or detail information of the working status data of the LED light-emitting component; the number of decomposition layers indicates the number of repeated decompositions. The more decomposition layers there are, the narrower the frequency band of the independent time series signal is divided and the higher the resolution is.

[0023] S42, collect the high-frequency detail coefficients obtained by wavelet decomposition of N signal channels at each layer, and construct a high-frequency detail transformation matrix for joint analysis between channels; the high-frequency detail transformation matrix is ;in, For the first signal channel The high-frequency detail coefficient matrix obtained by wavelet decomposition of the layer; For the The signal channel is The high-frequency detail coefficient matrix obtained by wavelet decomposition of the layer; For the The signal channel is The high-frequency detail coefficient matrix obtained by wavelet decomposition of the layer; S43, calculating the covariance matrix of the high-frequency detail coefficient matrix of each layer, which is used to describe the correlation between the high-frequency coefficients; ;in, For the The number of high-frequency detail coefficients in the layer high-frequency detail coefficient matrix; is the high frequency detail coefficient matrix The transposed matrix of is the transposition operator; the covariance matrix of each layer of high-frequency detail coefficient matrix obtained Perform eigenvalue decomposition; the covariance matrix after eigenvalue decomposition is ;in, is the covariance matrix after eigenvalue decomposition; is the covariance matrix The eigenvector matrix of is the covariance matrix The eigenvalue matrix is ​​usually a diagonal matrix, with the corresponding eigenvalues ​​on the diagonal, arranged from large to small; is the covariance matrix The transposed matrix of the eigenvector matrix is ​​used to reconstruct the result of eigendecomposition back to the original space; S44, using soft threshold processing method, the eigenvalue matrix Perform thresholding; preset threshold , the eigenvalue matrix The feature value is less than the preset threshold Set the part of to zero and get the new eigenvalue matrix ; The threshold processing formula is: ;in, After thresholding eigenvalues; is a symbolic function, representing the eigenvalue The positive and negative ; To keep The characteristic value of The eigenvalues ​​of are set to zero; is the preset threshold, which is obtained through the threshold adjustment formula; the threshold adjustment formula is ;in is the standard deviation of the noise; log is the sign of the logarithmic function; S45, through the new eigenvalue matrix , get the denoised covariance matrix ; Based on the denoised covariance matrix , use the reconstruction formula to calculate the high-frequency detail coefficient matrix after denoising; the reconstruction formula is ;in, For Take the square root of each term; S46, collecting low-frequency approximate coefficients obtained by wavelet decomposition of N signal channels at each layer, constructing a low-frequency approximate transformation matrix for joint analysis between channels, repeating steps S42-S45, and then obtaining a denoised low-frequency approximate coefficient matrix; S47. Combining and reconstructing the denoised low-frequency approximate coefficient matrix and the high-frequency detail coefficient matrix through inverse wavelet transform, thereby obtaining a working status feature data set.

[0024] The training methods for the policy optimization model include: The data set is divided into a training set, a validation set, and a test set; the data set includes a historical standardized digital control signal and a working state feature data set and a corresponding driver IC adjustment signal; a strategy optimization model is constructed, and the strategy optimization model includes an input layer, a hidden layer, and an output layer; the hidden layer uses a ReLU activation function; the input layer of the model is used to input the historical standardized digital control signal and the working state feature data set; the output layer of the model is used to output the driver IC adjustment signal, and the predicted value is output through a neuron; the strategy optimization model is a fully connected neural network model; Use mean absolute error as the loss function to measure the error between the model's predicted value and the actual value; Use the training set data to train the model and minimize the loss function using the Adam optimizer. Use the validation set to evaluate the performance of the model and tune the model's hyperparameters until the model performance stops improving or reaches the preset stopping condition. The test set is used to evaluate the performance of the model in the prediction task. The current standardized digital control signal and working state feature data set are input into the trained strategy optimization model to obtain the driver IC adjustment signal.

[0025] The method for adjusting the working state of the LED light emitting component according to the driving IC adjustment signal includes: The driver IC predicts the driver IC regulation signal through the strategy optimization model, and the driver IC regulation signal includes the current, voltage and PWM signal regulation parameters required by the LED light-emitting component; the driver IC regulation signal is converted into an actual output signal, and the working state of the LED light-emitting component is adjusted according to the actual output signal.

[0026] Methods for comparing the real-time light intensity output with the preset light intensity output to determine whether a deviation occurs include: If the real-time light intensity output is greater than or less than the preset light intensity output, it is determined that a deviation occurs; If the real-time light intensity output is equal to the preset light intensity output, it is determined that no deviation occurs; The preset light intensity output is set by the staff, and different light intensity outputs are collected through the driver IC control terminal, and the average value of multiple light intensity outputs is taken as the preset light intensity output; similarly, the minimum threshold of the preset light intensity deviation is set.

[0027] If a deviation occurs, the light intensity deviation is calculated, and the driver IC adjustment signal is further optimized based on the light intensity deviation until the light intensity output of the LED light emitting component reaches the optimal method, including: The driving IC adjustment signal is further optimized based on the light intensity deviation through the bacterial foraging optimization algorithm until the light intensity output of the LED light-emitting component reaches the optimal value; the specific method steps are as follows: S81. Calculate the light intensity deviation between the real-time light intensity output of the LED light-emitting component and the preset light intensity output using a light intensity deviation calculation formula; the light intensity deviation calculation formula is: ;in, The preset light intensity output; It is real-time light intensity output;

[0028] S82, further optimize the driver IC adjustment signal based on the light intensity deviation through the bacterial foraging optimization algorithm until the light intensity output of the LED light-emitting component reaches the optimal value; initialize the bacterial position, each bacterium represents a driver IC adjustment signal configuration; preset the population size to , each bacterial position is represented by a vector, denoted as ;in, For the Positional signals of bacteria; is the index of bacteria; the location of each bacterium Corresponding to a driver IC adjustment signal configuration, the real-time light intensity output of the LED is calculated through this configuration , and further calculate the light intensity deviation : ; S83. Define a fitness function to evaluate the quality of each bacterium to minimize the light intensity deviation; the fitness function is: ;in, For the The fitness value of each bacterium; S84, bacteria explore the solution space by swimming, and update their positions according to the current fitness value and the state of neighboring bacteria through the swimming update formula to find a better solution; the swimming update formula is: ;in, For the new locations of bacteria; For the the old location of the bacteria; is the step size factor that controls the swimming amplitude; is the global optimal position, that is, the driver IC adjustment signal configuration corresponding to the current optimal solution; For a between The random factor between S85. Dynamically adjust the step length factor for controlling the swimming amplitude through an adaptive step length formula; the adaptive step length formula is: ;in, is the dynamically adjusted step size factor; is the light intensity deviation of the current optimal solution; is the maximum light intensity deviation; A regulating factor for controlling the step size adjustment amplitude; S86. Limit the adjustment factor of the control step adjustment range by using the adjustment factor limiting formula; the adjustment factor limiting formula is: ;in, is the adjustment factor after restriction; is the number of current bacterial positions; The number of times the current bacterial position is updated; is the influence constant of the number of bacterial positions, which is used to balance the number of current bacterial positions the extent of the impact; is the influence constant of the bacterial position update, which is used to balance the number of times the current bacterial position is updated. the extent of the impact; The increase in the number of bacterial positions and the number of bacterial updates means that the search process gradually deepens and the accuracy of exploration gradually improves. Therefore, the step size should be gradually reduced as these parameters increase to avoid excessive step sizes that "skip" potential solution spaces in the later stages.

[0029] Each term in the formula ( and ) decreases with the increase of the number of bacteria and the number of updates, which represents the gradual convergence and refinement of the step size, which is consistent with the common “asymptotic convergence” characteristic in optimization algorithms.

[0030] In heuristic or swarm optimization algorithms, a larger step size is required in the exploration phase (early stage) to quickly traverse the solution space and avoid falling into the local optimal solution too early; while a smaller step size is required in the development phase (later stage) to search the solution space finely, improve accuracy, and avoid missing potential optimal solutions.

[0031] Through the formula ( and ), the step length is larger in the early stage (when and This dynamic adjustment is consistent with the strategy of balancing exploration and development in the optimization process.

[0032] The initial step size is large, allowing bacteria to move quickly in the solution space and avoid premature convergence. For example, when there are few bacterial locations ( Small), few updates, large step size, and can explore the solution space extensively.

[0033] Development stage: As evolution progresses, the step size decreases, and bacteria gradually concentrate in better areas, finding more accurate solutions through small step size adjustments. When it increases to a certain extent, the step length will be ( ) is reduced, making the search more precise.

[0034] constant and The influence of the number of bacteria and the number of updates on the step length is controlled. This allows the algorithm to flexibly adjust the rate at which the step length changes with the number of bacteria and the number of updates, and can be optimized according to the needs of actual problems.

[0035] For example, if you want the step size to decrease more quickly as the number of bacteria increases, you can increase ; If you want the step size to decrease faster as the number of updates increases, increase This provides freedom for adjusting the optimization algorithm and ensures that it can adapt to different environments and needs.

[0036] along with and As the value of increases, each term in the formula will gradually approach 0, making the step factor Gradually approaches 0. This naturally limits the size of the step size, ensuring that the step size will not be too large in the later stages of the algorithm, thereby ensuring the precision of the search. As the number of iterations increases or the search process deepens, the step size is gradually reduced to improve the accuracy of the local search.

[0037] For example, the adjustment factor that controls the step size adjustment is 0.8, the current number of bacterial positions is 30, the current number of bacterial position updates is 20, and the number of bacterial positions affects the constant is 0.02, the bacterial position update influence constant is 0.01, then the adjustment factor after restriction is .

[0038] S87. Use the random walk formula to simulate the random walk of bacteria in an unknown area; the random walk formula is: ;in, is the random walk step factor, which is used to control the jumping amplitude of bacteria in space; It is the optimal position for the bacterial colony; S88. After updating the position of each bacterium, recalculate its fitness, that is, calculate the new light intensity deviation, and evaluate its new fitness value; update the current optimal solution by comparing the fitness values; if the fitness of a certain bacterium is better, that is, the light intensity deviation is smaller, then update the position and fitness of the optimal solution; preset the minimum threshold of the light intensity deviation, when the light intensity deviation is The algorithm is stopped when the minimum threshold of the light intensity deviation is reached; the final optimized driver IC adjustment signal configuration is output, that is, the adjustment signal corresponding to the minimum light intensity deviation, at which time the light intensity output of the LED light-emitting component reaches the optimal level.

[0039] A method for controlling an LED light emitting component driver IC, comprising: A signal receiving module, used for receiving an external control signal; A signal processing module, used for preprocessing the received external control signal to obtain a standardized digital control signal; The data acquisition and processing module is used to collect the working state data of the LED light emitting component; pre-process the working state data of the LED light emitting component to obtain a working state characteristic data set; The optimization strategy module trains and acquires the strategy optimization model based on the standardized digital control signal and the working state characteristic data set, and predicts the driver IC adjustment signal; according to the driver IC adjustment signal, the working state of the LED light-emitting component is adjusted; The feedback adjustment module monitors the real-time light intensity output of the LED light-emitting component through a light sensor, compares the real-time light intensity output with the preset light intensity output, and determines whether a deviation occurs; if a deviation occurs, the light intensity deviation amount is calculated, and the driver IC adjustment signal is further optimized based on the light intensity deviation amount until the light intensity output of the LED light-emitting component reaches the optimal level.

[0040] The driver IC is located on the LED light-emitting component, and has a built-in circuit for implementing a method for controlling the driver IC of the LED light-emitting component. The driver IC can receive a driver IC adjustment signal from a control system of the driver IC of the LED light-emitting component, convert the driver IC adjustment signal into an actual output signal, and adjust the working state of the LED light-emitting component according to the actual output signal; when the actual output signal is executed, a method for controlling the driver IC of the LED light-emitting component is implemented.

[0041] This embodiment divides the working state data of the LED light-emitting component into multiple signal channels according to the data type, so that different types of data can be processed independently, avoiding interference between data types; through wavelet decomposition, the main trend (low-frequency approximate coefficient) and fast-changing or detail information (high-frequency detail coefficient) of the working state data of the LED light-emitting component can be accurately extracted, which is helpful for a deeper understanding of data characteristics; multi-channel joint analysis is performed using covariance matrix and eigenvalue decomposition to capture the cross-channel correlation between signal channels, thereby realizing multi-channel joint denoising; compared with single-channel denoising, it is more effective and can more accurately retain useful information and remove noise; the soft threshold processing method is used to threshold the eigenvalue matrix, which can effectively remove noise features while retaining important features, thereby improving the signal-to-noise ratio of the data; the denoised low-frequency approximate coefficient matrix and the high-frequency detail coefficient matrix are combined and reconstructed by inverse wavelet transform, so that a clearer and more accurate working state feature data set can be restored; By using the bacterial foraging optimization algorithm and simulating the swimming and foraging behavior of bacteria in the solution space, the configuration of the driver IC regulation signal can be explored and optimized quickly and efficiently; an adaptive step size formula is introduced, which can dynamically adjust the size of the step size factor according to the light intensity deviation of the current optimal solution, the maximum light intensity deviation and the regulation factor that controls the step size adjustment amplitude; it is helpful to quickly traverse the solution space in the early stage of the search to avoid falling into the local optimal solution too early; and in the later stage of the search, the step size can be gradually reduced to improve the accuracy of the local search; through the regulation factor restriction formula and the random walk formula, the method can balance the needs of the two stages of exploration and development; by continuously optimizing the configuration of the driver IC regulation signal, the light intensity output of the LED light-emitting component can be accurately controlled to reach the preset optimal value; by optimizing the configuration of the driver IC regulation signal, the light intensity deviation of the LED light-emitting component can be reduced, and the stability and reliability of the system can be improved; it is helpful to extend the service life of the LED light-emitting component, reduce the maintenance cost of the system, and improve the performance of the overall system.

[0042] Example 2 See also Figure 2 As shown, the part not described in detail in this embodiment can be seen from the description of embodiment 1, which provides an LED light emitting component driver IC control system, including: A signal receiving module, used for receiving an external control signal; A signal processing module, used for preprocessing the received external control signal to obtain a standardized digital control signal; The data acquisition and processing module is used to collect the working state data of the LED light emitting component; pre-process the working state data of the LED light emitting component to obtain a working state characteristic data set; The optimization strategy module trains and acquires the strategy optimization model based on the standardized digital control signal and the working state characteristic data set, and predicts the driver IC adjustment signal; according to the driver IC adjustment signal, the working state of the LED light-emitting component is adjusted; The feedback adjustment module monitors the real-time light intensity output of the LED light-emitting component through a light sensor, compares the real-time light intensity output with the preset light intensity output, and determines whether a deviation occurs; if a deviation occurs, the light intensity deviation amount is calculated, and the driver IC adjustment signal is further optimized based on the light intensity deviation amount until the light intensity output of the LED light-emitting component reaches the optimal level.

[0043] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the present invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0044] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the units is only one, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0045] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

[0046] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for controlling an LED light emitting component driver IC, characterized in that: include: S1, receiving external control signal; S2, preprocessing the received external control signal to obtain a standardized digital control signal; S3, collecting the working state data of the LED light emitting component; preprocessing the working state data of the LED light emitting component to obtain a working state feature data set; S4, training an acquisition strategy optimization model based on standardized digital control signals and working state feature data sets to predict the driver IC regulation signal; According to the driver IC adjustment signal, the working state of the LED light-emitting component is adjusted; S5. Monitor the real-time light intensity output of the LED light-emitting component through a light sensor, compare the real-time light intensity output with the preset light intensity output, and determine whether a deviation occurs; if a deviation occurs, calculate the light intensity deviation amount, and further optimize the driver IC adjustment signal based on the light intensity deviation amount until the light intensity output of the LED light-emitting component reaches the optimal level.

2. The LED light emitting component driver IC control method according to claim 1, characterized in that: The external control signals include brightness control signals, color temperature control signals, working mode control signals, current control signals, voltage control signals, environment monitoring signals and safety control signals; The working status data of the LED light-emitting component includes electrical parameter data, temperature parameter data, optical parameter data and operating time; the electrical parameter data includes current and voltage; the temperature parameter data includes the LED light-emitting component temperature and the driver IC temperature; the optical parameter data includes brightness and color temperature.

3. The LED light emitting component driver IC control method according to claim 2, characterized in that: The method of preprocessing the received external control signal to obtain a standardized digital control signal includes: The received external control signal is converted into a digital signal using an analog-to-digital converter, and is normalized to a uniform digital range by performing standardization processing; the normalized digital signal is denoised using a low-pass filter to obtain a denoised digital signal; the denoised digital signal is decoded into an operating mode instruction, and the obtained operating mode instruction is mapped into control parameters to obtain a standardized digital control signal.

4. The LED light emitting component driver IC control method according to claim 3, characterized in that: The method of preprocessing the working state data of the LED light-emitting component to obtain a working state feature data set includes: The working state data of the LED light-emitting component is divided into N signal channels according to different data types, and each signal channel is decomposed by wavelet to obtain low-frequency approximate coefficients and high-frequency detail coefficients; the working state data of the LED light-emitting component of the N signal channels are jointly analyzed through covariance matrix and eigenvalue decomposition, the cross-channel correlation between the N signal channels is captured, and multi-channel joint denoising is performed; the denoised low-frequency approximate coefficient matrix and high-frequency detail coefficient matrix are combined and reconstructed through inverse wavelet transform, and then the working state feature data set is obtained; the specific method steps are as follows: S41, dividing the working state data of the LED light-emitting component into N signal channels according to different data types, each signal channel containing a group of independent time series signals; performing timestamp alignment and normalization processing on the independent time series signals contained in the N signal channels, and integrating them into a time series vector containing the N signal channels; for the independent time series signals contained in each signal channel, using discrete wavelet transform to perform wavelet decomposition, and decompose them layer by layer into low-frequency approximate coefficients and high-frequency detail coefficients; S42, collect the high-frequency detail coefficients obtained by wavelet decomposition of N signal channels at each layer, and construct a high-frequency detail transformation matrix for joint analysis between channels; the high-frequency detail transformation matrix is ;in, For the first signal channel The high-frequency detail coefficient matrix obtained by wavelet decomposition of the layer; For the The signal channel is The high-frequency detail coefficient matrix obtained by wavelet decomposition of the layer; For the The signal channel is The high-frequency detail coefficient matrix obtained by wavelet decomposition of the layer; S43, calculate the covariance matrix of each layer of high-frequency detail coefficient matrix ;in, For the The number of high-frequency detail coefficients in the layer high-frequency detail coefficient matrix; is the high frequency detail coefficient matrix The transposed matrix of is the transposition operator; the covariance matrix of each layer of high-frequency detail coefficient matrix obtained Perform eigenvalue decomposition; the covariance matrix after eigenvalue decomposition is ;in, is the covariance matrix after eigenvalue decomposition; is the covariance matrix The eigenvector matrix of is the covariance matrix The eigenvalue matrix of is the covariance matrix The transposed matrix of the eigenvector matrix of ; S44, using soft threshold processing method, the eigenvalue matrix Perform thresholding; preset threshold , the eigenvalue matrix The feature value is less than the preset threshold Set the part of to zero and get the new eigenvalue matrix ; The threshold processing formula is: ;in, After thresholding eigenvalues; is a symbolic function, representing the eigenvalue The positive and negative ; To keep The characteristic value of The eigenvalues ​​of are set to zero; is the preset threshold, which is obtained through the threshold adjustment formula; the threshold adjustment formula is ;in is the standard deviation of the noise; log is the sign of the logarithmic function; S45, through the new eigenvalue matrix , get the denoised covariance matrix ; Based on the denoised covariance matrix , use the reconstruction formula to calculate the high-frequency detail coefficient matrix after denoising; the reconstruction formula is ;in, For Take the square root of each term; S46, collecting low-frequency approximate coefficients obtained by wavelet decomposition of N signal channels at each layer, constructing a low-frequency approximate transformation matrix for joint analysis between channels, repeating steps S42-S45, and then obtaining a denoised low-frequency approximate coefficient matrix; S47. Combining and reconstructing the denoised low-frequency approximate coefficient matrix and the high-frequency detail coefficient matrix through inverse wavelet transform, thereby obtaining a working status feature data set.

5. The LED light emitting component driver IC control method according to claim 4, characterized in that: The training method of the strategy optimization model includes: The data set is divided into a training set, a validation set and a test set; the data set includes a historical standardized digital control signal and a working state feature data set and a corresponding driver IC adjustment signal; a strategy optimization model is constructed, and the strategy optimization model includes an input layer, a hidden layer and an output layer; the hidden layer uses a ReLU activation function; the input layer of the model is used to input the historical standardized digital control signal and the working state feature data set; the output layer of the model is used to output the driver IC adjustment signal, and the predicted value is output through a neuron; the strategy optimization model is a fully connected neural network model; Use mean absolute error as the loss function to measure the error between the model's predicted value and the actual value; Use the training set data to train the model and minimize the loss function using the Adam optimizer. Use the validation set to evaluate the performance of the model and tune the model's hyperparameters until the model performance stops improving or reaches the preset stopping condition. The test set is used to evaluate the performance of the model in the prediction task. The current standardized digital control signal and working state feature data set are input into the trained strategy optimization model to obtain the driver IC adjustment signal.

6. The LED light emitting component driver IC control method according to claim 5, characterized in that: The method for adjusting the working state of the LED light emitting component according to the driving IC adjustment signal includes: The driver IC predicts the driver IC regulation signal through the strategy optimization model, and the driver IC regulation signal includes the current, voltage and PWM signal regulation parameters required by the LED light-emitting component; the driver IC regulation signal is converted into an actual output signal, and the working state of the LED light-emitting component is adjusted according to the actual output signal.

7. The LED light emitting component driver IC control method according to claim 6, characterized in that: The method of comparing the real-time light intensity output with the preset light intensity output to determine whether a deviation occurs includes: If the real-time light intensity output is greater than or less than the preset light intensity output, it is determined that a deviation occurs; If the real-time light intensity output is equal to the preset light intensity output, it is determined that no deviation occurs.

8. The LED light emitting component driver IC control method according to claim 7, characterized in that: If a deviation occurs, the light intensity deviation is calculated, and the driver IC adjustment signal is further optimized based on the light intensity deviation until the light intensity output of the LED light emitting component reaches the optimal value. The method includes: The driving IC adjustment signal is further optimized based on the light intensity deviation through the bacterial foraging optimization algorithm until the light intensity output of the LED light-emitting component reaches the optimal value; the specific method steps are as follows: S81. Calculate the light intensity deviation between the real-time light intensity output of the LED light-emitting component and the preset light intensity output using a light intensity deviation calculation formula; the light intensity deviation calculation formula is: ;in, The preset light intensity output; It is real-time light intensity output; S82, further optimize the driver IC adjustment signal based on the light intensity deviation through the bacterial foraging optimization algorithm until the light intensity output of the LED light-emitting component reaches the optimal value; initialize the bacterial position, each bacterium represents a driver IC adjustment signal configuration; preset the population size to , each bacterial position is represented by a vector, denoted as ;in, For the The position signal of each bacterium; is the index of bacteria; the location of each bacterium Corresponding to a driver IC adjustment signal configuration, the real-time light intensity output of the LED is calculated through this configuration , and further calculate the light intensity deviation : ; S83. Define a fitness function to evaluate the quality of each bacterium to minimize the light intensity deviation; the fitness function is: ;in, For the The fitness value of each bacterium; S84, bacteria explore the solution space by swimming, and update their positions according to the current fitness value and the state of neighboring bacteria through the swimming update formula to find a better solution; the swimming update formula is: ;in, For the new locations of bacteria; For the the old location of the bacteria; is the step size factor that controls the swimming amplitude; is the global optimal position, that is, the driver IC adjustment signal configuration corresponding to the current optimal solution; For a between The random factor between S85. Dynamically adjust the step length factor for controlling the swimming amplitude through an adaptive step length formula; the adaptive step length formula is: ;in, is the dynamically adjusted step size factor; is the light intensity deviation of the current optimal solution; is the maximum light intensity deviation; A regulating factor for controlling the step size adjustment amplitude; S86. Limit the adjustment factor of the control step adjustment range by using an adjustment factor limiting formula; the adjustment factor limiting formula is: ;in, is the adjustment factor after restriction; is the number of current bacterial positions; The number of times the current bacterial position is updated; is the constant affecting the number of bacterial positions; Update influence constants for bacterial positions; S87. Use the random walk formula to simulate the random walk of bacteria in an unknown area; the random walk formula is: ;in, is the random walk step factor, which is used to control the jumping amplitude of bacteria in space; It is the optimal position for the bacterial colony; S88. After updating the position of each bacterium, recalculate its fitness, that is, calculate the new light intensity deviation, and evaluate its new fitness value; update the current optimal solution by comparing the fitness values; if the fitness of a certain bacterium is better, that is, the light intensity deviation is smaller, then update the position and fitness of the optimal solution; preset the minimum threshold of the light intensity deviation, when the light intensity deviation is The algorithm is stopped when the minimum threshold of the light intensity deviation is reached; the final optimized driver IC adjustment signal configuration is output, that is, the adjustment signal corresponding to the minimum light intensity deviation, at which time the light intensity output of the LED light-emitting component reaches the optimal level.

9. An LED light emitting component driver IC control system, used to implement the LED light emitting component driver IC control method according to any one of claims 1 to 8, characterized in that: include: A signal receiving module, used for receiving an external control signal; A signal processing module, used for preprocessing the received external control signal to obtain a standardized digital control signal; A data acquisition and processing module is used to collect working status data of the LED light-emitting component; Preprocessing the working state data of the LED light-emitting component to obtain a working state feature data set; The optimization strategy module trains and acquires the strategy optimization model based on the standardized digital control signal and the working state characteristic data set, and predicts the driver IC adjustment signal; according to the driver IC adjustment signal, the working state of the LED light-emitting component is adjusted; The feedback adjustment module monitors the real-time light intensity output of the LED light-emitting component through a light sensor, compares the real-time light intensity output with the preset light intensity output, and determines whether a deviation occurs; if a deviation occurs, the light intensity deviation amount is calculated, and the driver IC adjustment signal is further optimized based on the light intensity deviation amount until the light intensity output of the LED light-emitting component reaches the optimal level.

10. A driver IC, characterized in that: The driver IC is located on the LED light-emitting component, and has a built-in circuit for implementing a method for controlling a driver IC of an LED light-emitting component. The driver IC can receive a driver IC adjustment signal from a control system of a driver IC of an LED light-emitting component, convert the driver IC adjustment signal into an actual output signal, and adjust the working state of the LED light-emitting component according to the actual output signal; when the actual output signal is executed, the method for controlling the driver IC of an LED light-emitting component described in any one of claims 1 to 8 is implemented.

Images