Full-color waveguide display method, system, medium and equipment based on liquid crystal polarizer grating and optoelectronic device

Abstract

The invention relates to the technical field of waveguide display, in particular to a full-color waveguide display method, system, medium and equipment based on a liquid crystal polarizer grating and a photoelectronic device, and the method comprises the steps: training and discriminating path direction information through a generative adversarial network, and meanwhile, setting an abnormal amplitude node based on initial polarization amplitude setting and abnormal amplitude node regulation and control; a directional channel cluster with amplitude constraint is formed, directional consistency and intensity continuity of path aggregation are realized, and through channel load offset ratio extraction, stable block identification and correction current mapping processing, light guide channels with different response characteristics obtain refined driving intensity control in a signal excitation process, so that the light guide effect is improved. In combination with an input channel signal configuration mode of a decision tree algorithm, dynamic closed-loop linkage between a signal control state and a path output result is realized by constructing a multi-attribute classification path, setting output frequency and displacement amplitude parameters, executing excitation signal bit width matching and synchronously loading frequency parameters.

Description

Technical Field

[0001] The present invention relates to the technical field of waveguide display, and particularly to a full-color waveguide display method, system, medium and device based on a liquid crystal polarization grating and optoelectronic devices. Background Art

[0002] The aim of the waveguide display technology field is to provide a display technology with high efficiency, flexibility and high-quality image performance through the characteristics of optical waveguides. It aims to guide optical signals through an optical waveguide system and precisely control the transmission path of light waves to achieve clear and vivid image display in a display device.

[0003] The aim of the full-color waveguide display method based on a liquid crystal polarization grating and optoelectronic devices is to modulate the polarization of light waves by using a liquid crystal polarization grating and combine the adjustment and control of light waves by optoelectronic devices, so as to achieve a full-color display effect in a waveguide system. The goal is to provide an efficient and stable display technology that can maintain excellent color reproduction and image quality at different angles and under different lighting conditions.

[0004] The prior art lacks the ability to dynamically model the path of different light propagation trajectories during the optical signal guiding process, resulting in a lack of real-time and adaptive mechanisms for path pointing and light intensity matching when facing changes in different angles or lighting conditions, and is prone to problems such as color shift or image blurring. When there are local fluctuations in the polarization state of some regions, it is difficult to quickly identify and process them at the fine-grained channel level, resulting in uneven brightness or obvious fluctuations in the edge region of the entire image. Moreover, most of the signal response paths are fixed connection structures, lacking a mapping regulation mechanism based on the characteristics of the current electrical signals. When the frequency response is unstable or the current distribution is uneven, path compensation and dynamic adjustment cannot be quickly performed, restricting the reliability expansion in a high-precision optical waveguide display system. Summary of the Invention

[0005] The purpose of the present invention is to solve the disadvantages existing in the prior art. In the first aspect, a full-color waveguide display method, system, medium and device based on a liquid crystal polarization grating and optoelectronic devices are provided.

[0006] To achieve the above object, the present invention adopts the following technical solutions: A full-color waveguide display method based on a liquid crystal polarization grating and optoelectronic devices, comprising the following steps:

[0007] Based on the pixel coordinate matrix distribution of an image display unit, perform a biaxial alignment setting on the polarization unit, generate a direction vector and extract trajectory points by rotating at a set angle and performing displacement translation, normalize the displacement difference between the trajectory points and screen stable paths, pair the propagation trajectories of the waveguide panels, and generate a pixel-level optical interference path set;

[0008] Based on the pixel-level optical interference path set, a generative adversarial network is adopted to extract path vectors and construct a partition map. By screening the same-direction components to aggregate path blocks, an initial polarization amplitude is set and the polarization amplitude threshold nodes are adjusted. Supplementary regulation is performed on the path anomaly change blocks, and the polarization rotation amplitude control nodes are updated by region to generate a polarization rotation amplitude regulation map;

[0009] Based on the polarization rotation amplitude regulation map, an optoelectronic device light source output excitation mechanism is constructed. By extracting the channel load offset ratio and sorting by partition, stable blocks and offset blocks are divided. A stable excitation signal is directly loaded, the excitation current of the offset blocks is corrected, and the channels are remapped to generate a pixel point excitation signal configuration table;

[0010] Based on the pixel point excitation signal configuration table, a decision tree algorithm is adopted. The signal combination is loaded into the optical waveguide channel. By synchronously setting the output frequency and displacement amplitude, the signal frequency displacement change difference is extracted. Tolerance interval combinations are sorted by error integral screening, and the channels are instructed to output and the signal states are recorded to generate a dynamic waveguide output configuration sequence.

[0011] As a further solution of the present invention, generating the pixel-level optical interference path set includes:

[0012] Based on the pixel coordinate matrix distribution of the image display unit, the biaxial position and angle of the polarization unit are corrected, and precise calibration of the position and angle rotation of each pixel point is performed. The angle and position are adjusted in sequence to ensure the precise registration of the polarization unit and the image coordinates, and a polarization direction vector is generated;

[0013] Based on the polarization direction vector, the position sequence nodes of each polarization unit are extracted, the displacement difference between the nodes is calculated, a normalization operation is performed to adjust the difference, and paths with high displacement difference stability and small deviation are screened to generate a stable path set;

[0014] Based on the stable path set, path matching calculation of the waveguide panel is performed on the propagation direction of the path, the propagation position of each path is calibrated and matched with the nodes to generate a pixel-level optical interference path set.

[0015] As a further solution of the present invention, generating the polarization rotation amplitude regulation map includes:

[0016] Based on the pixel-level optical interference path set, path vectors are extracted and divided according to the direction of the path. Sorting and classification of the path and direction are performed, a path area division structure is established, and the paths are distinguished by region to generate a path partition map;

[0017] Based on the above path partition map, use a generative adversarial network to screen path groups with the same direction and classify them into the same block. Perform path amplitude change stability evaluation and set the initial polarization amplitude. Adjust the control node state according to the polarization amplitude to generate a polarization amplitude regulation block;

[0018] Based on the polarization amplitude regulation block, regulate the regions with sudden changes in polarization amplitude in the path, perform the update calculation of the polarization rotation amplitude, and allocate corresponding control nodes according to the regions to generate a polarization rotation amplitude regulation map.

[0019] As a further solution of the present invention, the generative adversarial network follows the formula:

[0020] where: α represents the weight coefficient of the real sample evaluation loss, β represents the weight coefficient of the generated sample evaluation loss, x represents the input of the real path node set from the path partition map, P i data i represents the probability distribution of the real path sample, D represents the discriminator neural network structure, μ r represents the path polarization mean, σ r represents the path stability standard deviation, z represents the latent variable vector, which follows a normal distribution, P z z i represents the prior probability distribution of the latent variable, G represents the generator neural network structure, Θ c represents the path connection tightness coefficient, λ c represents the node heat adjustment factor, L i adv * represents the adversarial loss function value after adding conditional constraints.

[0021] As a further solution of the present invention, generating the pixel point excitation signal configuration table includes:

[0022] Based on the polarization rotation amplitude regulation map, extract the electrical load offset ratio of each pixel channel, perform pixel channel index traversal operations, extract the electrical signal amplitude sequence, calculate the ratio difference between the peak and mean of each channel, perform channel offset ratio mapping coding processing, reorder the channel indices in ascending numerical order, locate the channel partition threshold point, divide the regional boundaries of the channels and establish a load offset partition data set;

[0023] Based on the load offset partition data set, screen the channel groups with low electrical load change amplitude and directly load the excitation voltage with a fixed amplitude. Perform data bit addressing in the stable region, set the bit identification register structure for writing channel signals, write a constant current value, calculate the current correction coefficient for the offset region channels and synchronously update the control instructions to generate a set of corrected excitation current parameters;

[0024] Based on the corrected excitation current parameter set, perform index mapping and sorting on the excitation current of each channel, mark the pixel point positions corresponding to the channels, perform level classification processing on the channel excitation levels, call back the pixel coordinates corresponding to the channels, pair and number them and write them into the excitation data structure, and establish a pixel point excitation signal configuration table.

[0025] As a further solution of the present invention, generating the dynamic waveguide output configuration sequence includes:

[0026] Based on the pixel point excitation signal configuration table, using the decision tree algorithm, perform signal loading operations on the optical waveguide channels, synchronously set the output frequency and displacement amplitude parameters of each channel, perform bit width alignment processing of the excitation signals, read and write operations of the frequency parameters, and synchronous rising processing of the channel enable signals to generate a combined input channel control set;

[0027] Based on the combined input channel control set, collect the output frequency and displacement change data in real time, perform continuous data sampling, difference calculation and frequency shift sequence sorting of each channel, construct a channel error vector set, perform difference comparison calculation of two groups of signals and construct an error integral sequence, and establish a frequency shift difference integral sequence set;

[0028] Based on the frequency shift difference integral sequence set, perform error tolerance interval screening and parameter combination matching analysis, perform number classification processing of the error vector interval, maintain the index mapping table of the combined parameters, obtain the activation instruction and calibrate the corresponding channel number, drive the channel output and record each output state, and output the dynamic waveguide output configuration sequence.

[0029] As a further solution of the present invention, the decision tree algorithm is calculated according to the formula:

[0030]

[0031] Where: represents the Gini index of the excitation channel after multi-parameter weighting, K represents the total number of optical waveguide channels to be divided in the current excitation signal configuration task, ω i represents the synchronization priority coefficient of the i-th optical waveguide channel, p i represents the sample proportion of successful bit width matching of the excitation signal in the i-th channel, δ i represents the output frequency stability factor of the i-th channel, φ i represents the signal loading delay sensitivity factor of the i-th channel.

[0032] In a second aspect, a full-color waveguide display system based on a liquid crystal polarizer grating and an optoelectronic device is provided for performing any one of the full-color waveguide display methods based on a liquid crystal polarizer grating and an optoelectronic device in the first aspect. The system includes:

[0033] Pixel Trajectory Construction Module: Based on the pixel coordinate matrix distribution of the image display unit, set the biaxial azimuth angle parameters of the polarization unit, calculate the sequence of direction vectors formed by its rotation angle and displacement amount, extract the path node coordinates of the vectors, normalize the displacement difference between nodes and filter out the path set with a small change amplitude of the displacement difference, match the propagation direction data of the waveguide panel, and generate an optical interference path combination;

[0034] Polarization Amplitude Regulation Module: Based on the optical interference path combination, extract path vectors and construct a regional division graph structure, introduce a generative adversarial network to construct discriminant and generative sub-networks, filter out the path set with the same direction as the stable path, extract the product of the node amplitude and the number of the stable path, identify the control nodes exceeding the amplitude threshold, correct the amplitude abnormal data points, update the polarization rotation control data, and form a polarization regulation atlas set;

[0035] Channel Excitation Generation Module: Based on the polarization regulation atlas set, extract the rotation amplitude and excitation voltage, perform a weighted synthesis operation to generate an excitation source sequence, read the current load offset value of the channel and perform a variance value sorting process, divide the load stable area and the fluctuation area, load a constant excitation signal, adjust the channel mapping relationship of the fluctuation area and reconstruct the current excitation parameters, and output a pixel excitation parameter configuration table;

[0036] Waveguide Output Configuration Module: Based on the pixel excitation parameter configuration table, inject the excitation signal into the optical waveguide channel topology structure, extract the frequency and displacement amplitude of the output signal, calculate the difference between the two and perform sorting, use a decision tree algorithm to filter out the parameter combination segment where the difference falls within the tolerance range, label the channel signal and issue an output instruction, record the signal state data, and generate a dynamic output instruction linked list.

[0037] In a third aspect, a computer device is provided, including a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the full-color waveguide display system based on a liquid crystal polarization grating and an optoelectronic device as described in any item of the first aspect is implemented.

[0038] In a fourth aspect, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the full-color waveguide display method based on a liquid crystal polarization grating and an optoelectronic device as described in any item of the first aspect are implemented.

[0039] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0040] Train and discriminate the path direction information through a generative adversarial network. At the same time, based on the initial polarization amplitude setting and abnormal amplitude node regulation, form a direction channel cluster with amplitude constraints to achieve the direction consistency and intensity continuity of path aggregation. Through the extraction of the channel load offset ratio, the identification of stable blocks, and the correction of current mapping processing, realize the differential excitation configuration of the channels, so that the optical waveguide channels with different response characteristics obtain refined drive intensity control during the signal excitation process. Combine the input channel signal configuration method of the decision tree algorithm. By constructing a multi-attribute classification path, setting the output frequency and displacement amplitude parameters, and performing the matching of the excitation signal bit width and the synchronous loading of the frequency parameters, make each excitation decision process based on the signal feature classification and the recognition of the frequency shift rule, enhance the strategy consistency and instruction matching accuracy of the channel response, and realize the dynamic closed-loop linkage between the signal control state and the path output result. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 FIG. is a flowchart of a full-color waveguide display method based on a liquid crystal polarizer grating and an optoelectronic device shown according to an exemplary embodiment.

[0042] Figure 2 FIG. is a block diagram of a full-color waveguide display system based on a liquid crystal polarizer grating and an optoelectronic device shown according to an exemplary embodiment.

[0043] Figure 3 FIG. is a block diagram of a device for performing a full-color waveguide display method based on a liquid crystal polarizer grating and an optoelectronic device shown according to an exemplary embodiment. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0044] In order to make the objectives, technical solutions, and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0045] Please refer to Figure 1 , and a full-color waveguide display method based on a liquid crystal polarizer grating and an optoelectronic device is provided, including the following steps:

[0046] S1: Based on the pixel coordinate matrix distribution of the image display unit, perform a biaxial alignment setting on the polarization unit. By setting the angle rotation and performing displacement translation, generate a direction vector and extract trajectory points, normalize the displacement difference between the trajectory points, and screen stable paths, pair the waveguide panel propagation trajectories, and generate a pixel-level optical interference path set;

[0047] S2: Based on the pixel-level optical interference path set, using a generative adversarial network, extract path vectors and construct a partition map. Aggregate path blocks by screening co-directional components, set the initial polarization amplitude, and adjust the polarization amplitude threshold nodes. Perform supplementary regulation on path anomaly change blocks, update the polarization rotation amplitude control nodes by region, and generate a polarization rotation amplitude regulation map;

[0048] S3: Based on the polarization rotation amplitude regulation map, construct a light source output excitation mechanism for optoelectronic devices. By extracting the channel load offset ratio and sorting by region, divide into stable blocks and offset blocks, directly load stable excitation signals, correct the excitation current for offset blocks and remap the channels, and generate a pixel point excitation signal configuration table;

[0049] S4: Based on the pixel point excitation signal configuration table, use a decision tree algorithm, load signal combinations into the optical waveguide channel, by synchronously setting the output frequency and displacement amplitude, extract the signal frequency displacement change difference, sort by error integral to screen tolerance interval combinations, instruct the channel to output and record the signal state, and generate a dynamic waveguide output configuration sequence.

[0050] The pixel coordinate matrix distribution represents the arrangement structure of each pixel in the image display unit on a two-dimensional plane. The polarization unit represents an optical component used to control the polarization state of light waves. The direction vector represents the vector quantization result of the moving direction and angle of the pixel coordinates. The trajectory point represents the discrete path points formed after setting rotation and displacement operations. The normalized displacement difference between trajectory points represents the standardized processing result of the displacement change between trajectory points, used to eliminate scale effects. The optical interference path set represents the set of propagation and interference paths of light waves within the pixel level range. The path vector represents a vector group used to characterize the direction and intensity of the optical path. The partition map represents dividing the entire path set into several functional area layers according to clustering logic. The polarization amplitude represents the amplitude value of the electric field vector in the polarization state. The polarization amplitude threshold node represents the set threshold position for controlling the polarization adjustment range. The polarization rotation amplitude regulation map represents the regional mapping control map of the polarization angle change amount in the entire system. The light source output excitation mechanism represents the signal triggering mechanism for controlling the light emission process of optoelectronic devices. The channel load offset ratio represents the offset ratio between the signal load and the standard load in each optical waveguide channel. The excitation signal configuration table represents the set of light source excitation parameters allocated to each pixel. The decision tree algorithm represents a classification method for determining signal combination selection and channel configuration. The optical waveguide channel represents the waveguide path for transmitting controlled optical signals. The output frequency and displacement amplitude represent the change frequency and position movement amount of the signal during output. The error integral sorting represents the priority sorting by integrating the signal error magnitude. The tolerance interval combination represents the acceptable signal configuration combination within a certain error range. The dynamic waveguide output configuration sequence represents the set of instruction sequences for controlling the dynamic response of the optical waveguide channel.

[0051] In step S1, the specific steps for generating the pixel-level optical interference path set are as follows:

[0052] Based on the pixel coordinate matrix distribution of the image display unit, perform biaxial position and angle correction on the polarization unit, execute precise calibration of the position and angle rotation of each pixel point, adjust the angle and position in sequence to ensure accurate registration of the polarization unit with the image coordinates, and generate a polarization direction vector;

[0053] Based on the polarization direction vector, extract the position sequence nodes of each polarization unit, calculate the displacement differences between the nodes, perform a normalization operation to adjust the differences, filter out the paths with high displacement difference stability and small deviation, and generate a stable path set;

[0054] Based on the stable path set, perform path matching calculation of the waveguide panel for the propagation direction of the paths, calibrate the propagation position of each path and match it with the nodes, and generate a pixel-level optical interference path set;

[0055] The pixel coordinate matrix distribution represents the geometric arrangement structure of all pixels in the image display unit on a two-dimensional plane. The polarization unit represents an optical component used to control the polarization direction of light waves. The biaxial position and angle correction represents the precise adjustment of the position and rotation angle of the polarization unit in the horizontal and vertical directions. The polarization direction vector represents the vector expression form of the direction information of the polarization unit after registration at each pixel. The position sequence node represents the spatial position points of the polarization units arranged in pixel order. The displacement difference represents the physical position difference between adjacent position sequence nodes. The normalization operation represents the standardization process of the displacement differences to eliminate the interference caused by inconsistent scales. The stable path represents a set of pixel connection paths with small offset errors and high position consistency. The stable path set represents the combined set after filtering out all stable paths. The waveguide panel represents an integrated optical structure for transmitting optical signals. The path matching calculation represents the calculation process of spatially mapping and aligning the propagation direction of the stable paths with the transmissible paths in the waveguide structure. The propagation position represents the actual transmission spatial position of the path in the waveguide structure. The pixel-level optical interference path set represents a set of light propagation and interference paths established based on each pixel and matched with the waveguide;

[0056] Based on the pixel coordinate matrix distribution of the image display unit, an affine transformation method is used to perform affine transformation on the biaxial alignment of the polarization unit. By constructing a rotation matrix and a translation matrix to perform two-dimensional coordinate transformation, the input parameters of the rotation matrix include the rotation angle around the X-axis of 0.05π, the rotation angle around the Y-axis of 0.025π, and the rotation angle around the Z-axis of 0.01π. The input parameters of the translation matrix include an offset of 2 microns in the X direction and an offset of 3 microns in the Y direction. The execution order is to perform the rotation matrix transformation first and then the translation transformation. Each point in the pixel coordinate matrix is sequentially transformed to complete the angle and position calibration operation of the pixel points, and the adjusted position coordinate data is output. The angle and position are adjusted sequentially to generate a polarization direction vector matrix;

[0057] Based on the polarization direction vector matrix, an interpolation algorithm for filling points of the polarization unit trajectory points is adopted. By extracting the displacement difference between adjacent trajectory points for vector linear interpolation processing, the linear interpolation parameters include the starting vector, the ending vector, and the number of interpolation steps of 5 steps. The interpolation method is equidistant interval linear accumulation calculation. Subsequently, a normalization operation is performed using the maximum and minimum normalization method. The input parameters include the normalization target interval of 0 to 1, and the execution method is to perform scaling conversion according to the maximum and minimum values of the difference sequence. After the difference normalization, the path points are screened. The screening criterion is that the normalized value is less than 0.1. The double average filtering algorithm is used to process the path point sequence. The length of the first-stage filtering window is 5 points. The average value calculation within the moving window is performed and the mean difference comparison is carried out. The mean difference less than 0.02 is the retention criterion. In the second stage, the variance within the window is statistically calculated, and the variance less than 0.005 is the consistency determination criterion. Only the path points that meet the double conditions are retained to generate a stable path set;

[0058] Based on the stable path set, a waveguide pairing algorithm based on plane vector normal fitting is used to perform waveguide panel pairing on the path propagation direction. By extracting the path direction vector and comparing it with the waveguide surface normal vector for direction consistency calculation, the direction consistency determination criterion is that the vector angle is less than 5 degrees. The vector angle calculation method is inner product comparison. After matching the path points with consistent directions, a propagation position calibration operation is performed. The pixel-level Bresenham integer stepping method is used for path discrete mapping processing. The starting point coordinates and the ending point coordinates are input, and the step-by-step integer stepping path calculation is performed and the coordinate values of each pixel point during the stepping process are sequentially recorded. The position binding and mapping confirmation are performed for each path point and the discrete coordinate point to generate a pixel-level optical interference path set.

[0059] In step S2, the specific steps for generating the polarization rotation amplitude control map are as follows:

[0060] Based on the pixel-level optical interference path set, extract path vectors and divide them according to the direction of the paths. Perform sorting and classification of the paths and directions, establish a path region division structure, distinguish the paths by region, and generate a path partition map;

[0061] Based on the path partition map, use a generative adversarial network to screen out path groups with consistent directions and classify them into the same block. Perform an evaluation of the stability of the path amplitude change and set the initial polarization amplitude. Adjust the state of the control nodes according to the polarization amplitude to generate a polarization amplitude regulation block;

[0062] Based on the polarization amplitude regulation block, regulate the regions with sudden changes in polarization amplitude in the paths, perform updated calculations of the polarization rotation amplitude, and allocate corresponding control nodes according to the regions to generate a polarization rotation amplitude regulation map;

[0063] The pixel-level optical interference path set represents the set of light wave interference propagation paths established at the single-pixel scale. The path vector represents the vector data used to describe the path direction and geometric features. The path direction represents the directional information of light propagation between pixels. The path region division structure represents the structural organization method of dividing paths with similar direction and position characteristics into multiple physical or functional regions. The path partition map represents the regional division diagram result generated according to the path direction and spatial distribution. The path group with consistent directions represents the set of paths with consistency or high similarity in the propagation direction. The evaluation of the stability of the path amplitude change represents the stability analysis of the polarization amplitude fluctuation degree of the path optical signal. The initial polarization amplitude represents the initial amplitude value of the polarization electric field set during the path regulation process. The control node state represents the operating state of the control points used to adjust the polarization amplitude in the electrical or optical control system. The polarization amplitude regulation block represents the functional block identified as needing to adjust the polarization amplitude in the path partition. The region with sudden change in polarization amplitude represents the position of the polarization amplitude where mutation or transition occurs in the path. The polarization rotation amplitude represents the angular range of the polarization vector rotation in space. The polarization rotation amplitude regulation map represents the mapping control diagram structure for controlling the polarization rotation degree at the regional level;

[0064] Based on the pixel-level optical interference path set, a quicksort algorithm based on path direction vector partitioning is used to perform direction partitioning on the path vectors. By extracting the two-dimensional direction vectors formed by the starting and ending points of each path, the overall direction classification of the paths is calculated using the direction angle θ. In the calculation of the direction angle, the polar angle reference is set as the horizontal axis, and the direction angle interval is divided into 0 to 360 degrees, divided into one category every 15 degrees, and a total of 24 direction intervals are set. After the direction partitioning, sorting processing is performed. The quicksort method is used for sorting, and the path direction angle values are compared as the primary key, and ascending sorting is set. After the sorting is completed, the path position attribution block annotation operation is executed. The corresponding partition index is established based on the divided direction range, and the quadrant where the path coordinate end point is located is used for auxiliary confirmation to construct a mapping relationship table between the paths and the blocks. The block mapping numbers are completed for all paths and a two-dimensional annotation layer is generated to generate a path partition map;

[0065] Based on the path partition map, the generative adversarial network method is used to perform automatic merging operations on the co-directional paths. The network architecture used is the standard GAN structure. Among them, the input of the generator is the direction category and amplitude distribution matrix of the path vector, the input tensor size is 1×64, the activation function is ReLU, the number of hidden layer nodes is 128, and the output is the path attribution category label. The input of the discriminator is the concatenated data of the path vector and the corresponding category label, the input dimension is 1×128, the activation function is LeakyReLU, and the discriminant result is a prediction scalar of whether it belongs to the same category. The Adam optimizer is used to train the parameters, where the learning rate is set to 0.0002, β1 is 0.5, and the number of training rounds is 400 rounds. The co-directional paths are screened out and labeled with a unified category number, and the path stability evaluation operation is executed. The sliding window mean square error method is used to calculate the path direction change. The window width is 7 path units, and the mean square error lower than 0.01 is the stable path judgment threshold. The initial polarization amplitude is set to 0.75 in the stable paths. The node state parameters on each path are adjusted according to the amplitude value. The node current modulation parameter is set to 1.2 mA, and the initial voltage excitation value is 2.3 V. The adjustment method is to directly map the amplitude to the current amplitude change ratio, and the unified synchronization time step is 10 milliseconds to generate a polarization amplitude regulation block;

[0066] Based on the polarization amplitude regulation block, the amplitude outlier detection and vector rotation compensation algorithm are used to regulate the blocks with abnormal amplitude changes in the path. The abnormal amplitude judgment criterion is that the amplitude deviates from the block average value by more than ±20%. In the detection process, the amplitude value of each node is processed by moving average filtering, the window length is 9. After filtering, the difference between each node and the average value is calculated, and the amplitude rotation compensation operation is performed. The rotation vector adjustment mechanism is used to approximate the polarization angle from the current value to the adjacent stable angle in steps of 0.1 degree, and the maximum number of iteration steps is 100. If the step size exceeds the limit, the current state is retained and the abnormality is recorded. In the node regulation process, the master node-driven broadcast mechanism within the region is adopted. One master node is set for each block, and the broadcast frequency is 20Hz. The regulation signal format is 16-bit binary polarization amplitude encoding. The signal reception range is set to all child nodes within a radius of 30 pixels. After decoding the received signal, the polarization angle and amplitude value are adjusted to generate a polarization rotation amplitude regulation map.

[0067] Generative adversarial network, according to the formula:

[0068]

[0069] where: α represents the weight coefficient of the evaluation loss of the real sample, β represents the weight coefficient of the evaluation loss of the generated sample, x represents the input of the real path node set from the path partition map, P i data i represents the probability distribution of the real path sample, D represents the discriminator neural network structure, μ r represents the path polarization mean, σ r represents the path stability standard deviation, z represents the latent variable vector, which follows a normal distribution, P z z i represents the prior probability distribution of the latent variable, G represents the generator neural network structure, Θ c represents the path connection tightness coefficient, λ c represents the node heat adjustment factor, L i adv * represents the adversarial loss function value after adding conditional constraints;

[0070] Execution process: First, extract the channel number to be configured and the corresponding excitation signal bit width standard from the pixel point excitation signal configuration table as the input features of the decision tree. Then, obtain the channel frequency upper limit and the minimum resolution step size according to the preset frequency regulation mapping table to construct the channel output frequency node feature set. Next, calculate the bit width ratio p i of the matched signals in each channel, and substitute it into the Gini index formula to calculate I G(t), the priority channel for the loading operation to be performed under the current conditions is determined by recursive partitioning. Subsequently, the frequency value writing, bit width configuration synchronization, and enable bit pulling high operations are completed in the order of priority. Finally, a combined input channel control set is generated, which includes the excitation signal number, target channel index, frequency write value, bit width configuration index, and enable flag bit field, constituting a unified configuration sequence for all channels.

[0071] In step S3, the specific steps for generating the pixel excitation signal configuration table are as follows:

[0072] Based on the polarization rotation amplitude regulation map, the electrical load offset ratio of each pixel channel is extracted, the pixel channel index traversal operation is performed, the electrical signal amplitude sequence is extracted, the ratio difference between the peak and mean values of each channel is calculated, the channel offset ratio mapping coding process is performed, the channel index is reordered in ascending numerical order, the channel partition threshold point is located, the regional boundary of the channel is divided, and a load offset partition data set is established;

[0073] Based on the load offset partition data set, the channel group with a low electrical load change amplitude is screened and a fixed amplitude excitation voltage is directly loaded. The data bit addressing of the stable region is performed, the bit identification register structure for writing the channel signal is set, a constant current value is written, the current correction coefficient is calculated for the offset region channels, and the control instruction is synchronously updated to generate a corrected excitation current parameter set;

[0074] Based on the corrected excitation current parameter set, index mapping and sorting are performed on the excitation current of each channel, the pixel point position corresponding to the channel is marked, the level classification process of the channel excitation level is performed, the pixel coordinates corresponding to the channel are called back, paired numbers are written into the excitation data structure, and a pixel excitation signal configuration table is established;

[0075] The polarization rotation amplitude regulation map represents an image-based control data structure for controlling the magnitude of the polarization rotation angle in each region. The electrical load offset ratio represents the relative offset between the actual electrical load and the standard load in the pixel channel. The pixel channel index represents the numbering system used to identify each pixel electrical channel. The electrical signal amplitude sequence represents the sequence of electrical signal intensity values collected in the pixel channel as a function of time or position. The ratio difference between the peak and the mean represents the relative deviation between the peak and the average of the electrical signal in the channel, which is used to measure the degree of signal fluctuation. The offset ratio mapping and encoding process represents the conversion of the electrical load offset ratio value into structured numerical data through function mapping and encoding. The ascending order sorting of the channel index represents the reordering of the channel numbers from smallest to largest according to the mapping and encoding values. The channel partition threshold point represents the numerical node used to divide the boundary of the channel region. The load offset partition data set represents the data set obtained by dividing the channels according to the load offset characteristics. The electrical load change amplitude represents the change amount of the electrical signal in the channel in the time or space dimension. The excitation voltage represents the voltage signal used to drive the pixel to emit light or respond to an action. The data bit addressing represents the address identification operation for determining the storage location to write the control signal. The bit identification register structure represents the register data structure used to identify and store the signal writing location. The constant current value represents the fixed drive current that does not change with time or operation. The current correction coefficient represents the proportional correction parameter required to adjust the current. The control instruction represents the operation command used to adjust the behavior of the hardware. The corrected excitation current parameter set represents the set of current configuration parameters formed after the correction calculation. The index mapping represents the establishment of the correspondence between the channel number and the excitation parameter. The excitation level classification process represents the grouping process of the current signal according to the intensity interval. The excitation data structure represents the data organization form storing the relationship between the excitation parameter, the channel number, and the pixel coordinates. The pixel point excitation signal configuration table represents the configuration list of the electrical signal excitation parameters assigned to each pixel;

[0076] Based on the polarization rotation amplitude regulation map, the pixel channel offset coding method based on signal amplitude ratio difference analysis is used to extract the offset ratio values of each pixel channel load. By traversing each channel and performing pixel number index scanning, the channel ID is called in sequence, and the corresponding electrical signal amplitude sequence of the channel is extracted from the digital signal input buffer. The amplitude sequence contains 128 sampling points, and the sampling period of each point is 10 microseconds. Calculate the difference between the maximum value and the mean value of the electrical signal sequence of each channel to form a channel peak-to-average difference set. Use the ratio calculation method to construct an offset ratio matrix, set the ratio calculation accuracy to three significant digits after the decimal point, perform the offset ratio mapping coding algorithm on all channels for numbering, use the sequential coding method to assign numbers in ascending order of the offset ratio from 0 to N, sort the generated coding sequence in ascending order, and the sorting algorithm uses quicksort. The input parameter is the ratio vector sequence. After sorting, perform partition boundary threshold positioning. Set the sliding detection window size to 7, and use the condition that the mean mutation rate of adjacent windows is greater than 25% as the threshold judgment basis. Mark each boundary point that meets the threshold condition as the start of region division, perform channel number region segmentation, and establish a load offset partition data set;

[0077] Based on the load offset partition data set, the fixed voltage excitation loading and block addressing mechanism is used to process the load stable blocks. By detecting the blocks in the offset block with a standard deviation of ratio change less than 0.02, they are identified as stable blocks, and a fixed amplitude excitation voltage is directly loaded. The excitation voltage value is set to 2.8 volts, and the continuous loading time is 80 milliseconds. The loading method uses a constant voltage source to directly access the corresponding channel output control unit. Perform the stable block data bit addressing operation, use the address mapping table index mechanism based on the linear storage structure, set the address step to 4-bit binary units, assign a write bit identifier to each channel, and the identifier format is an 8-bit mask. The first 4 bits are the channel region number, and the last 4 bits are the channel body serial number. The operation of writing a fixed current value uses a constant current loading algorithm, and the loading current value is 5 milliamps, and the holding time is set to 100 milliseconds. Calculate the current correction factor for all channels in the offset block using the difference inverse deduction correction factor calculation method. The input parameters are the standard current value and the current channel offset coefficient. The calculation method is to multiply the current value by the reciprocal of the offset factor and then multiply by the standard current. All correction factors correspond one-to-one with the original channel mapping numbers, and are constructed and synchronously updated to the control instruction data buffer to generate a corrected excitation current parameter set;

[0078] Based on the corrected excitation current parameter set, an index mapping operation is performed on the excitation current of each channel using the combined algorithm of channel current index rearrangement and position mapping function. The combined mapping table is established by calling the channel excitation current sequence and the channel number list. The index values in the mapping table are used as the primary key for ascending rearrangement. The channel labeling rule is to bind the physical pixel position coordinate information corresponding to the excitation current value of each channel. The binding structure is a triple structure containing the channel ID, excitation current value, and pixel coordinates. The channel excitation level classification operation is performed, and the classification threshold is segmented at intervals of 0.1 mA. Each segment is independently assigned a level label. The channel pixel coordinate callback operation obtains the corresponding pixel point by querying the channel-pixel mapping table, and calls the structured query function to match the ID and coordinate fields for coordinate write-back. The paired number writing structure uses the nested key-value mapping storage format generation function. The number pair format is that the channel ID and level value form the key, and the corresponding value is the written excitation value, and a pixel point excitation signal configuration table is established.

[0079] In step S4, the specific steps for generating the dynamic waveguide output configuration sequence are as follows:

[0080] Based on the pixel point excitation signal configuration table, using the decision tree algorithm, signal loading operations are performed on the optical waveguide channels, the output frequencies and displacement amplitude parameters of each channel are synchronously set, the bit width alignment processing of the excitation signal, the read and write operations of the frequency parameters, and the synchronous rising processing of the channel enable signal are performed to generate a combined input channel control set;

[0081] Based on the combined input channel control set, the output frequency and displacement change data are collected in real time, continuous data sampling, difference calculation, and frequency shift sequence sorting are performed on each channel to construct a channel error vector set, the difference comparison calculation of two groups of signals is performed and an error integral sequence is constructed to establish a frequency shift difference integral sequence set;

[0082] Based on the frequency shift difference integral sequence set, error tolerance interval screening and parameter combination matching analysis are performed, number classification processing of the error vector interval and maintenance operation of the index mapping table of the combination parameters are carried out to obtain the activation instruction and calibrate the corresponding channel number, drive the channel output and record each output state, and output the dynamic waveguide output configuration sequence;

[0083] The pixel excitation signal configuration table is represented as a set of signal configuration data preset for each pixel, including excitation parameters such as current, voltage, and frequency. The decision tree algorithm represents a machine learning method for path branch selection based on feature conditions, used to determine channel control operations. The optical waveguide channel represents the optical waveguide path structure for transmitting the excitation signal to the target pixel. The output frequency represents the number of signal changes per unit time. The displacement amplitude parameter represents the physical movement amount of the signal in space. The bit-width alignment process represents unifying the data bit-width of the excitation signal to the standard width required by the system. The frequency parameter read / write operation represents the process of accessing and modifying the set frequency values for each channel. The enable signal synchronous rising process represents the control logic for simultaneously activating multiple channels to achieve concurrent output. The combined input channel control set represents a unified control instruction set formed by integrating multiple input channel control parameters. The output frequency and displacement change data represent the dynamic change values of frequency and displacement collected by each channel during actual operation. Continuous data sampling represents continuously collecting the channel output data at fixed time intervals. The difference calculation represents the operation of measuring the numerical change between adjacent sampling points. The frequency shift sequence sorting represents the process of arranging the frequency displacement change amounts of each channel in ascending order. The channel error vector set represents a set of vectors recording the error between the output signal of each channel and the target reference signal. The error integral sequence represents the time series result after integrating and accumulating the channel error values. The frequency shift difference integral sequence set represents the set data summarizing the integral results of the frequency and displacement differences of each channel. The error tolerance interval represents the error range section acceptable to the system. The parameter combination matching analysis represents screening the combination schemes that meet the tolerance requirements from multiple groups of configuration parameters. The error vector interval number classification process represents the classification process of dividing the error data into intervals and assigning unique number labels. The index mapping table represents a data structure establishing the correspondence between the channel number and the parameter combination. The activation instruction represents a command used to start a specific channel to execute the output task. The channel number represents the unique identifier of each optical waveguide channel in the system. The output state record represents the state storage of the signal result generated by the channel after activation. The dynamic waveguide output configuration sequence represents a set of optical waveguide channel signal output schemes with dynamic adaptability generated according to the error regulation strategy;

[0084] Based on the pixel excitation signal configuration table, the decision tree algorithm is used to perform signal loading operations on the optical waveguide channels. By constructing a training sample set that includes attribute nodes "excitation bit width", "output frequency", "displacement amplitude", and "channel ID", the decision tree structure is constructed using the ID3 algorithm. The information gain threshold is set to 0.3, the attribute partitioning criterion is the principle of the smallest difference in integer values of the frequency, the maximum depth of the decision node is set to 5, the path is generated for the training samples and the class nodes are output. The class output is the selected signal loading method. The loading method includes three dimensions: bit width setting, level type, and waveform structure. The signal bit width matching uses an 8-bit alignment mechanism. When writing to each channel, the bit width is padded to the byte boundary. The frequency parameter is read synchronously. The frequency setting range is from 1 kHz to 20 kHz, the step value is 500 Hz, the reading method is to call the data in the corresponding index through the channel mapped address, the writing execution method is to directly call the register write function, the address segment is from 0x1000 to 0x10FF, the set value range of the displacement amplitude parameter is from 0 to 5 mm, the parsing step size is 0.2 mm, the channel enable bit is set by pulling up the GPIO control signal. The GPIO address mapping range starts from 0x2000, and each channel occupies one control bit. After exciting each channel, the write control register status is "1", and a combined input channel control set is generated;

[0085] Based on the combined input channel control set, a data stream acquisition method based on the DMA sampling buffer mechanism is used to collect output frequency and displacement change data. Each channel is continuously sampled at a sampling frequency of 1 MHz through a high-speed sampling interface. The sampling window length is 128 clock cycles. All sampling values are sequentially written into the data buffer FIFO. After sampling is completed, a difference calculation operation is performed. The difference algorithm is the absolute difference calculation between adjacent sampling points. The difference sequence is arranged in the order of sampling time. A frequency shift sequence is constructed. The length of each sequence is fixed at 64 groups of samples. An error vector is constructed for each channel sequence. The error vector is generated by comparing the current channel frequency shift sequence with the previous sequence. The comparison method is to accumulate the differences at the same index position to form a vector. The vector dimension is 64. The mean square error comparison method is used to compare the two groups of signal difference sequences. The maximum error limit is set to 1.5 Hz, and the minimum integration accuracy is 0.05 Hz. After calculating the error integral, it is written into the error sequence record structure. The structure fields include channel ID, sampling time, integral value, and difference amplitude level identifier. Finally, a frequency shift difference integral sequence set is established;

[0086] Based on the frequency shift difference integral sequence set, the error interval screening and matching operations are processed using the error interval mapping and tolerance mapping control table update algorithm. The error interval screening condition is that the integral difference is greater than 2 Hz and the duration exceeds 50 milliseconds. The channel numbers that meet the conditions are added to the screening list. The list element structure includes the channel number, start sampling time, end sampling time, and integral value level. The error vector interval number classification operation is performed. The classification method is divided into four categories according to the integral value interval range, corresponding to different output channel activation strategies. The operation of updating the combined parameter index mapping table calls the dynamic mapping update interface function. The interface input parameters include the interval number, tolerance range, and channel status code. The mapping table format is a two-dimensional hash table. The key-value pair consists of the interval number and the channel ID. The value field stores the current excitation intensity, activation delay, and duration. The generated activation instruction structure includes the target channel ID, activation type code, control duration, and status latch flag bit. The corresponding channel output signal is executed through the drive control pin. The drive signal is updated and generated by the controller according to the set value in the mapping table. Finally, all channel output status flags are recorded and summarized and encoded into a unified status table, and the dynamic waveguide output configuration sequence is output.

[0087] The decision tree algorithm, according to the formula:

[0088]

[0089] Where: represents the Gini index of the excitation channel after multi-parameter weighting, K represents the total number of optical waveguide channels to be divided in the current excitation signal configuration task, ω i represents the synchronization priority coefficient of the i-th optical waveguide channel, p i represents the sample proportion of successful bit width matching of the excitation signal in the i-th channel, δ i represents the output frequency stability factor of the i-th channel, φ i represents the signal loading delay sensitivity factor of the i-th channel;

[0090] Execution process: First, extract the target bit width, polarization direction, and target frequency of each optical waveguide channel from the liquid crystal polarizer grating control matrix, and calculate the bit width matching ratio p i , then based on the frequency sequence sampled at the output end of the optoelectronic device, calculate the channel output frequency stability factor δ i , obtained through the signal variance inverse model. Then, extract the average write delay cycle number from the response log of the control signal loaded on each channel, and construct the channel loading delay sensitivity factor φ i , and then combine the dependence degree of each channel in the scheduling control unit on the combined waveguide projection image area to determine the synchronization priority coefficient ω i , and calculate according to the parameters uniformly substituted into the weighted Gini index formula Determine whether the current optical waveguide channel is preferentially divided and excited, and finally generate a combined input channel control set that supports frequency control, polarization compensation, and displacement synchronization, so as to achieve precise excitation control of the image channels in each polarization direction of the full-color waveguide structure.

[0091] Please refer to Figure 2 , this embodiment of the present disclosure also provides a full-color waveguide display system based on a liquid crystal polarization grating and optoelectronic devices, which is used to execute the full-color waveguide display method based on a liquid crystal polarization grating and optoelectronic devices. The system includes:

[0092] A pixel trajectory construction module, configured to set the biaxial azimuth angle parameters of the polarization unit based on the pixel coordinate matrix distribution of the image display unit, calculate the direction vector sequence formed by its rotation angle and displacement amount, extract the path node coordinates of the vector, normalize the displacement difference between nodes, and screen the path set with a small change amplitude of the displacement difference, match the propagation direction data of the waveguide panel, and generate an optical interference path combination;

[0093] A polarization amplitude regulation module, configured to extract path vectors and construct a region division graph structure based on the optical interference path combination, introduce a generative adversarial network to construct discriminant and generative sub-networks, screen the path set with the same direction as a stable path, extract the product of the node amplitude and the number of the stable path, identify the control nodes exceeding the amplitude threshold, correct the amplitude abnormal data points, update the polarization rotation control data, and form a polarization regulation map set;

[0094] A channel excitation generation module, configured to extract the rotation amplitude and excitation voltage based on the polarization regulation map set, perform a weighted synthesis operation to generate an excitation source sequence, read the current load offset value of the channel and perform variance value sorting processing, divide the load stable region and the fluctuation region, load a constant excitation signal, adjust the channel mapping relationship of the fluctuation region and reconstruct the current excitation parameters, and output a pixel excitation parameter configuration table;

[0095] A waveguide output configuration module, based on the pixel excitation parameter configuration table, injects the excitation signal into the optical waveguide channel topology structure, extracts the frequency and displacement amplitude of the output signal, calculates the difference between the two and performs sorting, uses a decision tree algorithm, screens the parameter combination segments whose differences fall within the tolerance range, labels the channel signals and issues output instructions, records the signal status data, and generates a dynamic output instruction linked list.

[0096] In some possible implementation manners, the pixel trajectory construction module is configured to:

[0097] Based on the pixel coordinate matrix of the image display unit, perform biaxial alignment of the polarization unit and pixel position angle calibration, adjust the angle and position to ensure the precise alignment of the polarization unit and the coordinates, and generate a polarization direction vector;

[0098] Based on the polarization direction vector, extract trajectory points, calculate the displacement difference and perform normalization processing, screen for stable paths, and generate a set of stable paths;

[0099] Based on the set of stable paths, match the propagation direction of the waveguide panel, calibrate the path propagation position and complete path point matching, and generate a pixel-level optical interference path set.

[0100] In some possible implementation manners, the polarization amplitude regulation module is configured to:

[0101] Based on the pixel-level optical interference path set, extract path vectors and sort them by direction, establish path partitions and generate a path partition map;

[0102] Based on the path partition map, use a generative adversarial network to screen for co-directional paths, classify them into the same block, evaluate path stability and set an initial polarization amplitude, adjust node states, and generate a polarization amplitude regulation block;

[0103] Based on the polarization amplitude regulation block, regulate the amplitude abnormal block, update the polarization rotation amplitude, allocate regulation nodes by region, and generate a polarization rotation amplitude regulation map.

[0104] In some possible implementation manners, the polarization amplitude regulation module is configured to generate the adversarial network according to the following formula:

[0105]

[0106] Where: α represents the weight coefficient of the real sample evaluation loss, β represents the weight coefficient of the generated sample evaluation loss, x represents the input of the real path node set from the path partition map, P i data i represents the probability distribution of the real path sample, D represents the discriminator neural network structure, μ r represents the path polarization mean, σ r represents the path stability standard deviation, z represents the latent variable vector, which follows a normal distribution, P z z i represents the prior probability distribution of the latent variable, G represents the generator neural network structure, Θ c represents the path connection tightness coefficient, λ c represents the node heat regulation factor, L i adv * represents the adversarial loss function value after adding conditional constraints.

[0107] In some possible implementation manners, the channel excitation generation module is configured to:

[0108] Based on the polarization rotation amplitude regulation map, extract the load offset ratio of each pixel channel, traverse the channel numbers, extract the electrical signal amplitude sequence, calculate the ratio of the peak value to the mean value difference, perform offset ratio mapping coding, sort in ascending numerical order and locate the partition threshold points, divide the region boundaries, and establish a load offset partition data set;

[0109] Based on the load offset partition data set, screen the stable blocks to load a fixed amplitude excitation voltage, complete data bit addressing, set the electrical signal write bit identifier, write the constant current value, calculate the current correction factor for the offset blocks and synchronously update the control instructions to generate a corrected excitation current parameter set;

[0110] Based on the corrected excitation current parameter set, re-index the excitation current of each channel and label the corresponding pixel point positions, perform excitation level classification, coordinate callback, and pair number writing into the excitation data structure to establish a pixel point excitation signal configuration table.

[0111] Based on the pixel point excitation signal configuration table, use the decision tree algorithm to perform signal loading on the optical waveguide channels, set the output frequency and displacement amplitude, perform excitation signal bit width matching, frequency parameter reading and writing, and synchronously pull up the channel opening bit to generate a combined input channel control set;

[0112] Based on the combined input channel control set, collect the output frequency and displacement changes in real time, perform continuous reading sampling, difference calculation, and frequency shift sequence arrangement, construct a channel error vector set, calculate the difference between two groups of signals and generate an error integral sequence, and establish a frequency shift difference integral sequence set;

[0113] Based on the frequency shift difference integral sequence set, screen the error intervals and perform tolerance combination matching, complete the error vector interval number classification, update the combined parameter index mapping table, generate an activation instruction and point to the target channel, drive the channel output, record the output status, and output a dynamic waveguide configuration sequence.

[0114] The decision tree algorithm, according to the formula:

[0115]

[0116] Where: represents the Gini index of the excitation channel after multi-parameter weighting, K represents the total number of optical waveguide channels to be divided in the current excitation signal configuration task, ω i represents the synchronization priority coefficient of the i-th optical waveguide channel, p i represents the sample proportion of successful excitation signal bit width matching in the i-th channel, δ i represents the output frequency stability factor of the i-th channel, φ i represents the signal loading delay sensitivity factor of the i-th channel.

[0117] An embodiment of the present disclosure further provides a computer device, including a memory and a processor. A computer program is stored in the memory. When the processor executes the computer program, the full-color waveguide display system based on a liquid crystal polarizer grating and an optoelectronic device as described above is implemented.

[0118] An embodiment of the present disclosure further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the full-color waveguide display method based on a liquid crystal polarizer grating and an optoelectronic device as described above are implemented.

[0119] Figure 3 The device 100 for executing the full-color waveguide display method based on a liquid crystal polarizer grating and an optoelectronic device as shown includes: a processor 1001 and a memory 1003. Among them, the processor 1001 and the memory 1003 are connected, such as connected through a bus 1002. Optionally, the device 100 for executing the full-color waveguide display method based on a liquid crystal polarizer grating and an optoelectronic device may further include a communication component, and the communication component may be used for data interaction between the device 100 and other devices, such as sending and / or receiving data, etc. It should be noted that in actual scheduling, the communication component is not limited to one, and the structure of the device 100 for executing the full-color waveguide display method based on a liquid crystal polarizer grating and an optoelectronic device does not constitute a limitation to the embodiments of the present application.

[0120] The processor 1001 may be a CPU (Central Processing Unit, central processing unit), a general-purpose processor, a DSP (Digital Signal Processor, data signal processor), an ASIC (Application Specific Integrated Circuit, application-specific integrated circuit), an FPGA (Field Programmable Gate Array, field programmable gate array) or other programmable logic devices, transistor logic devices, hardware components or any combination thereof. It can implement or execute various exemplary logic blocks, modules and circuits described in combination with the disclosure of the present application. The processor 1001 may also be a combination for implementing computing functions, such as a combination including one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0121] The bus 1002 may include a path for transmitting information between the above components. The bus 1002 can be a PCI (Peripheral Component Interconnect) bus, an EISA (Extended Industry Standard Architecture) bus, or the like. The bus 1002 can be divided into an address bus, a data bus, a control bus, etc. For the sake of convenience of representation, Figure 3 it is only represented by a thick line in Figure 3 , but it does not mean that there is only one bus or one type of bus.

[0122] The memory 1003 can be a ROM (Read Only Memory), or other types of static storage devices that can store static information and instructions, a RAM (Random Access Memory), or other types of dynamic storage devices that can store information and instructions. It can also be an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read Only Memory), or other optical disc storage, optical disc storage (including compact discs, laser discs, optical discs, digital versatile discs, Blu-ray discs, etc.), magnetic disk storage media, other magnetic storage devices, or any other medium that can be used to carry or store program code and can be read by a computer, which is not limited herein.

[0123] The memory 1003 is used to store the program code for implementing the embodiments of the present disclosure, and is controlled by the processor 1001 to execute. The processor 1001 is used to execute the program code stored in the memory 1003 to implement the steps shown in the foregoing embodiments of the blower drive motor control method.

[0124] The embodiments of the present disclosure also provide a computer-readable storage medium, on which program code is stored. When the program code is executed by a processor, the steps and corresponding contents of the foregoing embodiments of the blower drive motor control method can be implemented.

[0125] The above are only the preferred embodiments of the present invention, and do not limit the present invention in other forms. Any person skilled in the art may use the disclosed technical content to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the technical solution content of the present invention still belong to the protection scope of the technical solution of the present invention.

[0126] The preferred embodiments of the present disclosure have been described in detail above in conjunction with the accompanying drawings. However, the present disclosure is not limited to the specific details in the above embodiments. Within the technical concept of the present disclosure, various changes, modifications, substitutions, and variations can be made to these embodiments, and these changes, modifications, substitutions, and variations all fall within the protection scope of the present disclosure.

[0127] In addition, it should be noted that, among the various specific technical features described in the above specific embodiments, they can be combined in any appropriate manner without contradiction, and the same should be regarded as the content disclosed by the present disclosure. To avoid unnecessary repetition, the present disclosure will not separately describe various possible combination methods. The technical scope of this application is not limited to the content in the specification, and its technical scope must be determined according to the scope of the claims.

Claims

1. A full-color waveguide display method based on a liquid crystal polarization grating and an optoelectronic device, characterized in that, It includes the following steps: Based on the pixel coordinate matrix distribution of the image display unit, perform biaxial alignment setting on the polarization unit. Generate a direction vector and extract trajectory points by rotating at a set angle and performing displacement translation. Normalize the displacement difference between trajectory points and screen stable paths, and pair the propagation trajectories of waveguide panels to generate a pixel-level optical interference path set; Based on the pixel-level optical interference path set, use a generative adversarial network to extract path vectors and construct a partition map. Aggregate path blocks by screening the same-direction components, set the initial polarization amplitude, adjust the polarization amplitude threshold nodes, perform supplementary regulation on path abnormally changing blocks, and update the polarization rotation amplitude control nodes by region to generate a polarization rotation amplitude regulation map; Based on the polarization rotation amplitude regulation map, construct an optoelectronic device light source output excitation mechanism. Divide stable blocks and offset blocks by extracting the channel load offset ratio and sorting by region. Directly load stable excitation signals, correct the excitation current for offset blocks and remap channels to generate a pixel point excitation signal configuration table; Based on the pixel point excitation signal configuration table, use a decision tree algorithm to load signal combinations into the optical waveguide channel. Extract the difference in signal frequency displacement by synchronously setting the output frequency and displacement amplitude, sort by error integral to screen tolerance interval combinations, instruct the channel to output and record the signal state to generate a dynamic waveguide output configuration sequence.

2. The full-color waveguide display method based on a liquid crystal polarizer grating and an optoelectronic device according to claim 1, wherein Generating the pixel-level optical interference path set includes: Based on the pixel coordinate matrix of the image display unit, perform biaxial alignment of the polarization unit and pixel position angle calibration, adjust the angle and position to ensure precise alignment of the polarization unit with the coordinates, and generate a polarization direction vector; Based on the polarization direction vector, extract trajectory points, calculate the displacement difference and perform normalization processing, screen stable paths, and generate a stable path set; Based on the stable path set, match the propagation direction of the waveguide panel, calibrate the path propagation position and complete path point matching to generate a pixel-level optical interference path set.

3. The full-color waveguide display method based on a liquid crystal polarization grating and an optoelectronic device according to claim 1, wherein Generating the polarization rotation amplitude regulation map includes: Based on the pixel-level optical interference path set, extract path vectors and sort them by direction, establish path partitions and generate a path partition map; Based on the path partition map, use a generative adversarial network to screen same-direction paths, classify them into the same block, evaluate path stability and set the initial polarization amplitude, adjust the node state, and generate a polarization amplitude regulation block; Based on the polarization amplitude regulation block, regulate the amplitude abnormally changing block, update the polarization rotation amplitude, allocate regulation nodes by region, and generate a polarization rotation amplitude regulation map.

4. The full-color waveguide display method based on a liquid crystal polarization grating and an optoelectronic device according to claim 3, wherein The generative adversarial network is calculated according to the formula: Where: α represents the weight coefficient of the real sample evaluation loss, β represents the weight coefficient of the generated sample evaluation loss, x represents the input of the real path node set from the path partition map, P i data i represents the probability distribution of the real path sample, D represents the discriminator neural network structure, μ r represents the path polarization mean, σ r represents the path stability standard deviation, z represents the latent variable vector, following a normal distribution, P z z i represents the prior probability distribution of the latent variable, G represents the generator neural network structure, Θ c represents the path connection tightness coefficient, λ c represents the node heat adjustment factor, L i adv * represents the adversarial loss function value after adding conditional constraints.

5. The full-color waveguide display method based on a liquid crystal polarization grating and an optoelectronic device according to claim 1, wherein Generating the pixel point excitation signal configuration table includes: Based on the polarization rotation amplitude regulation map, extract the load offset ratio of each pixel channel, traverse the channel numbers, extract the electrical signal amplitude sequence, calculate the ratio of the peak value to the mean value difference, perform offset ratio mapping coding, sort in ascending order of numerical value and locate the partition threshold point, divide the region boundary, and establish a load offset partition data set; Based on the load offset partition data set, filter the stable blocks to load a fixed amplitude excitation voltage, complete data bit addressing, set the electrical signal write bit identifier, write the constant current value, calculate the current correction factor for the offset blocks and synchronously update the control instructions, and generate a corrected excitation current parameter set; Based on the corrected excitation current parameter set, re-index the excitation current of each channel and label the corresponding pixel positions, perform excitation level classification, coordinate callback, and pair number writing into the excitation data structure, and establish a pixel excitation signal configuration table.

6. The full-color waveguide display method based on a liquid crystal polarization grating and an optoelectronic device according to claim 1, characterized in that Generate the dynamic waveguide output configuration sequence, including: Based on the pixel excitation signal configuration table, use the decision tree algorithm to perform signal loading on the optical waveguide channels, set the output frequency and displacement amplitude, perform excitation signal bit width matching, frequency parameter reading and writing, and synchronously pull up the channel opening bits, and generate a combined input channel control set; Based on the combined input channel control set, collect the output frequency and displacement changes in real time, perform continuous reading sampling, difference calculation and frequency shift sequence arrangement, construct a channel error vector set, calculate the difference between two groups of signals and generate an error integral sequence, and establish a frequency shift difference integral sequence set; Based on the frequency shift difference integral sequence set, filter the error interval and perform tolerance combination matching, complete the error vector interval number classification, update the combined parameter index mapping table, generate an activation instruction and point to the target channel, drive the channel output, record the output status, and output the dynamic waveguide output configuration sequence.

7. The full-color waveguide display method based on a liquid crystal polarization grating and an optoelectronic device according to claim 6, wherein The decision tree algorithm, according to the formula: Wherein: represents the Gini index of the excitation channel after multi-parameter weighting, K represents the total number of optical waveguide channels to be divided in the current excitation signal configuration task, ω i represents the synchronization priority coefficient of the i-th optical waveguide channel, p i represents the proportion of samples with successful matching of the excitation signal bit width in the i-th channel, δ i represents the output frequency stability factor of the i-th channel, φ i represents the signal loading delay sensitivity factor of the i-th channel.

8. A full-color waveguide display system based on a liquid crystal polarization grating and an optoelectronic device, characterized in that, The full-color waveguide display system based on the liquid crystal polarizer grating and optoelectronic devices is used to execute the full-color waveguide display method based on the liquid crystal polarizer grating and optoelectronic devices according to any one of claims 1 to 7. The system includes: Pixel trajectory construction module: Based on the pixel coordinate matrix of the image display unit, set the biaxial direction angles of the polarization unit, calculate the rotation angle and displacement amount, generate a direction vector sequence, extract the trajectory points, normalize the displacement difference of the trajectory points and filter the stable paths, and match the waveguide panel propagation direction data to generate an optical interference path combination; Polarization amplitude regulation module: Based on the optical interference path combination, construct a partition map, introduce a generative adversarial network for path screening, extract the amplitude and quantity product of the stable path block nodes, identify the polarization amplitude threshold nodes and correct the outliers, and update the polarization rotation data to form a polarization regulation map set; Channel excitation generation module: Based on the polarization regulation map set, extract the rotation amplitude and excitation voltage, generate an excitation source combination sequence, read the channel load offset current value, perform variance sorting, divide the stable and offset blocks, load the stable excitation signal, adjust the offset block mapping relationship, reconstruct the current excitation parameters, and output the pixel excitation parameter configuration table; Waveguide output configuration module: Based on the pixel excitation parameter configuration table, inject the excitation signal into the optical waveguide channel, extract the output signal frequency and displacement amplitude, calculate the difference and sort it, use the decision tree algorithm to screen the tolerance interval combination, mark the channel signal and command the output, record the signal status, and generate a dynamic output instruction linked list.

9. A computer device, comprising a memory and a processor, characterized in that, A computer program is stored in the memory, and when the processor executes the computer program, it implements the full-color waveguide display method based on a liquid crystal polarization grating and an optoelectronic device according to any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the full-color waveguide display method based on a liquid crystal polarization grating and an optoelectronic device according to any one of claims 1-7.

Images