LED module light intensity control method and system
By classifying the light intensity data of the LED module and processing the machine learning model, combined with the power state dynamic adjustment of the light intensity control instructions, the problem of insufficient flexibility and accuracy of light intensity control in the existing technology is solved, and the LED module is realized The precise and stable adjustment of the light intensity of the group meets the lighting needs in varying environments.
Patent Information
- Application Number
- CN202410994715.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-24
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2044-07-24
AI Technical Summary
The existing LED module light intensity control technology is difficult to achieve uniform and automatic optimization of light intensity under different working conditions. It is affected by the nonlinear light output characteristics and external factors of the LED lamp beads, resulting in insufficient control flexibility and accuracy.
By obtaining the initial light intensity data and state information of the LED module, performing light intensity type classification and data cleaning, using preset machine learning models for light intensity feature extraction and clustering analysis, dynamically adjusting the light intensity control instructions to adapt to the variable frequency power supply state, realize multimodal fusion and real-time adjustment of light intensity.
It improves the accuracy and stability of light intensity control, can provide ideal lighting effects in a variable working state, meet the needs of different scenarios, save energy consumption, and improve the energy efficiency performance of the system.
Smart Images

Figure CN118555701B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of data processing, and in particular to a method and system for controlling light intensity of an LED module. Background Art
[0002] In lighting and display technology, the use of LED modules is popular due to its advantages such as energy saving, long life and small size. The light intensity control of LED modules is of great significance in meeting the lighting requirements of different application scenarios, improving visual comfort and realizing intelligent lighting. This usually requires precise control and adjustment of the brightness of each LED lamp bead to achieve the final lighting effect. Therefore, how to effectively adjust and control the light intensity of LED modules has become an important research and development direction in this field.
[0003] At present, the existing LED module light intensity control method mainly achieves light intensity change by changing the power supply current driving the LED lamp beads or adjusting the duty cycle of the PWM (pulse width modulation) signal. The existing technology can meet the regulation needs to a certain extent, but it faces many challenges, such as how to ensure uniform light intensity under different working conditions, and how to comprehensively consider the influence of the LED lamp beads' own characteristics and their working environment on light intensity regulation.
[0004] In practical applications, existing LED module light intensity control technology is often subject to the influence of external factors such as the nonlinear light output characteristics of LED lamp beads and power supply changes, resulting in insufficient control flexibility and accuracy. In addition, when the LED module works under different lighting environments or load conditions, it is difficult for existing methods to automatically optimize the light intensity output to adapt to these changes. Therefore, a more accurate and intelligent light intensity control method is needed to optimize the light output characteristics and ensure a more ideal lighting effect under variable working conditions. Summary of the invention
[0005] The present invention provides a method and system for controlling light intensity of an LED module, which are used to solve the above-mentioned technical problems.
[0006] A first aspect of the present invention provides a method for controlling light intensity of an LED module, the method comprising:
[0007] Acquire initial light intensity data of the LED module to be controlled and initial state information of the corresponding LED lamp beads, and classify the initial light intensity data into light intensity types to obtain classified light intensity data; wherein the LED module to be controlled includes at least a plurality of LED lamp beads;
[0008] Matching brightness adjustment features corresponding to the classified light intensity data in a preset database to obtain respective brightness adjustment feature data;
[0009] Based on the preset light intensity type weight, each brightness adjustment feature data is subjected to adjustment feature multi-modal fusion to obtain a fusion result, and based on the fusion result, a light intensity control instruction corresponding to the initial light intensity data is generated;
[0010] Acquire current working state information of the variable frequency power supply of the LED module to be controlled, and dynamically adjust the light intensity control instruction based on the current working state information of the variable frequency power supply of the LED module to be controlled to obtain an adjusted light intensity control instruction;
[0011] Based on the adjusted light intensity control instruction and the initial state information of the LED lamp beads, the initial state of the LED lamp beads is adjusted to generate a light intensity control adjustment result, and based on the light intensity control adjustment result, the light intensity control adjustment is performed on the LED module to be controlled.
[0012] Optionally, in a first implementation of the first aspect of the present invention, the initial light intensity data of the LED module to be controlled and the initial state information of the corresponding LED lamp beads are obtained, and the initial light intensity data is classified into light intensity types to obtain the classified light intensity data, including:
[0013] Performing data cleaning on the initial light intensity data to obtain the initial light intensity data after data cleaning; wherein the initial light intensity data at least includes brightness data, color temperature data, and color saturation data;
[0014] Normalizing the initial light intensity data after data cleaning to obtain normalized initial light intensity data;
[0015] Performing statistical analysis on the normalized initial light intensity data to obtain distribution characteristics of the initial light intensity data; wherein the distribution characteristics of the initial light intensity data at least include mean, variance, skewness, kurtosis, maximum value, and minimum value;
[0016] Performing feature extraction on the normalized initial light intensity data to obtain first feature data and second feature data; wherein the first feature data at least includes average amplitude, standard deviation, and root mean square value; and the second feature data at least includes power spectrum density, frequency distribution, and harmonic components;
[0017] Performing cluster analysis on the distribution characteristics of the initial light intensity data, the first feature data, and the second feature data to obtain a plurality of initial light intensity features;
[0018] Smoothing the multiple initial light intensity features, eliminating short-term fluctuations in the multiple initial light intensity features, and obtaining smoothed light intensity feature data;
[0019] The smoothed light intensity feature data is classified into light intensity types using a preset light intensity type classification model to obtain classified light intensity data; wherein the preset light intensity type classification model is a machine learning model obtained through advance training.
[0020] Optionally, in a second implementation of the first aspect of the present invention, the training process of the machine learning model includes:
[0021] Collecting LED module light intensity training data under multiple different environmental conditions; wherein the LED module light intensity training data at least includes brightness training data, color temperature training data, and color saturation training data;
[0022] Obtaining a preset machine learning model; wherein the preset machine learning model includes at least an input layer, a multiple feature extraction layer, a feature embedding layer, a fusion layer, a multi-layer fully connected network layer, and a multi-level classification layer, and the multi-level classification layer includes at least a primary classifier and a refined classifier;
[0023] Inputting the LED module light intensity training data into the input layer for preprocessing to obtain the preprocessed LED module light intensity training data;
[0024] The pre-processed LED module light intensity training data is transmitted to the multiple feature extraction layers layer by layer to obtain a multi-dimensional feature vector;
[0025] The extracted multi-dimensional feature vector is input into the feature embedding layer, and the multi-dimensional feature vector is embedded in high dimension to obtain multiple high-dimensional embedding features;
[0026] The high-dimensional embedded features are input into the fusion layer, and each high-dimensional embedded feature is fused to obtain a comprehensive light intensity feature vector;
[0027] The comprehensive light intensity feature vector is input into a multi-layer fully connected network layer, and the comprehensive light intensity feature vector is extracted and combined layer by layer to obtain the light intensity feature representation;
[0028] Input the light intensity feature representation into the multi-level classification layer, perform preliminary classification on the light intensity feature representation through the primary classifier in the multi-level classification layer to obtain a preliminary classification result, and perform refinement classification on the preliminary classification result through the refinement classifier in the multi-level classification layer to obtain a light intensity type label;
[0029] During the training process, combined with the preset actual light intensity type label, the loss function of the model is continuously iteratively optimized through the preset back propagation algorithm, and the model parameters are adjusted until the loss function converges to obtain the trained machine learning model as the light intensity type classification model; wherein the light intensity type classification model is used to classify the smoothed light intensity feature data into light intensity types to obtain classified light intensity data.
[0030] Optionally, in a third implementation of the first aspect of the present invention, based on a preset light intensity type weight, performing adjustment feature multimodal fusion on each brightness adjustment feature data to obtain a fusion result, and based on the fusion result, generating a light intensity control instruction corresponding to the initial light intensity data, including:
[0031] Obtaining weight parameters of multiple light intensity types; wherein the weight parameters are set based on preset environmental requirements and light intensity adjustment targets, and stored in a database;
[0032] The weight parameters are used to assign weights to the brightness adjustment feature data, and the contribution of the feature values of the brightness adjustment feature data is adjusted according to the weights of different light intensity types to generate a weighted feature matrix;
[0033] The weighted feature matrix is fused to obtain a comprehensive adjustment feature vector as a fusion result; wherein the fusion process includes a data weighted average method and a modality cross fusion method;
[0034] The fusion result is input into a preset light intensity control instruction generation algorithm to generate a corresponding light intensity control instruction.
[0035] Optionally, in a fourth implementation of the first aspect of the present invention, after the step of inputting the fusion result into a preset light intensity control instruction generation algorithm to generate a corresponding light intensity control instruction, the step further includes:
[0036] According to the light intensity control instruction, the light intensity parameters of the LED module are controlled, and the initial light intensity data of the LED lamp beads are adjusted in real time;
[0037] In the real-time adjustment process, the current light intensity data of the LED module is monitored and collected based on the preset feedback loop, compared with the initial light intensity data, and the weight parameters and characteristic values of the brightness adjustment characteristic data are adjusted;
[0038] Based on the adjusted weight parameters and the adjusted characteristic values of the brightness adjustment characteristic data, the generated light intensity control instructions are iteratively adjusted and optimized to obtain the optimized light intensity control instructions, and the optimized light intensity control instructions are applied to the LED module.
[0039] Optionally, in a fifth implementation of the first aspect of the present invention, after the step of inputting the fusion result into a preset light intensity control instruction generation algorithm to generate a corresponding light intensity control instruction, the step further includes:
[0040] Convert the generated light intensity control instruction into a character string in a preset format to obtain a character string of the light intensity control instruction;
[0041] Dynamically generate encryption keys according to preset encryption key generation rules; wherein the generated encryption keys are regularly updated based on preset security policies and environmental requirements;
[0042] The character string of the light intensity control instruction is encrypted using the generated encryption key to obtain the encrypted light intensity control instruction;
[0043] The encrypted light intensity control instruction is stored in a database, and a corresponding verification code is generated; wherein the verification code is used to ensure the data integrity and correctness of the encrypted light intensity control instruction during the transmission process.
[0044] Optionally, in a sixth implementation of the first aspect of the present invention, after the step of storing the encrypted light intensity control instruction in a database and generating a corresponding verification code, the method further comprises:
[0045] The transmission path is encrypted by a preset secure communication protocol to obtain an encrypted transmission path; wherein the encrypted transmission path is used to securely transmit the encrypted light intensity control instruction and the corresponding verification code to the LED module to prevent it from being obtained or tampered with by unauthorized parties;
[0046] After the LED module receives the encrypted light intensity control instruction and the corresponding verification code, the LED module is verified based on the corresponding verification code. If the verification passes, the LED module is authorized to start the decryption program; wherein the decryption program includes: using a preset decryption key generation rule to dynamically generate a decryption key, decrypting the encrypted light intensity control instruction based on the generated decryption key, and obtaining the decrypted light intensity control instruction.
[0047] A second aspect of the present invention provides an LED module light intensity control system, the LED module light intensity control system comprising:
[0048] An acquisition module is used to acquire initial light intensity data of the LED module to be controlled and initial state information of the corresponding LED lamp beads, and classify the initial light intensity data into light intensity types to obtain classified light intensity data; wherein the LED module to be controlled includes at least a plurality of LED lamp beads;
[0049] A matching module, used to match the brightness adjustment features corresponding to the classified light intensity data in a preset database to obtain each brightness adjustment feature data;
[0050] A fusion module is used to perform multi-modal fusion of adjustment features on each brightness adjustment feature data based on a preset light intensity type weight to obtain a fusion result, and generate a light intensity control instruction corresponding to the initial light intensity data based on the fusion result;
[0051] An adjustment module, used for acquiring current working state information of the variable frequency power supply of the LED module to be controlled, and dynamically adjusting the light intensity control instruction based on the current working state information of the variable frequency power supply of the LED module to be controlled to obtain an adjusted light intensity control instruction;
[0052] The control and adjustment module is used to adjust the initial state of the LED lamp beads based on the adjusted light intensity control instructions and the initial state information of the LED lamp beads, generate a light intensity control adjustment result, and based on the light intensity control adjustment result, perform light intensity control adjustment on the LED module to be controlled.
[0053] The third aspect of the present invention provides an LED module light intensity control device, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory so that the LED module light intensity control device executes the above-mentioned LED module light intensity control method.
[0054] A fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the above-mentioned LED module light intensity control method.
[0055] The technical solution provided by the present invention has the following beneficial effects: the present invention provides a method and system for controlling light intensity of an LED module, which obtains initial light intensity data of an LED module to be controlled and corresponding initial state information of LED lamp beads, and classifies the initial light intensity data into light intensity types to obtain classified light intensity data; matches brightness adjustment features corresponding to the classified light intensity data in a preset database to obtain various brightness adjustment feature data; performs multimodal fusion of adjustment features on various brightness adjustment feature data based on preset light intensity type weights to obtain fusion results, and generates light intensity control instructions corresponding to the initial light intensity data based on the fusion results; obtains current working state information of a variable frequency power supply of an LED module to be controlled, and dynamically adjusts the light intensity control instructions based on the current working state information of the variable frequency power supply of the LED module to be controlled to obtain adjusted light intensity control instructions; adjusts the initial state of the LED lamp beads based on the adjusted light intensity control instructions and the initial state information of the LED lamp beads to generate a light intensity control adjustment result, and performs light intensity control adjustment on the LED module to be controlled based on the light intensity control adjustment result. The present invention obtains the initial state information of each LED lamp bead and its corresponding light intensity, and classifies these light intensity data, so that the light intensity control is more accurate, and appropriate control strategies can be formulated for different types of light intensity requirements. The brightness adjustment features associated with the classified light intensity data are matched in a preset database, and the most suitable brightness adjustment parameters can be selected according to the specific conditions of different LED modules, thereby optimizing the light output effect of the entire system. Based on the preset light intensity type weight, multimodal fusion can be performed, and a variety of adjustment features can be combined to achieve a comprehensive adjustment effect, which helps to alleviate the problems of uneven light output caused by a single adjustment parameter. The light intensity control instruction is dynamically adjusted according to the current working state information of the variable frequency power supply, so that the brightness adjustment of the LED module can change in real time according to the power supply state, thereby realizing efficient and flexible brightness adaptive control. The dynamic control process integrating the initial light intensity data, the power supply state information and the brightness adjustment feature can more effectively respond to changes in external conditions and ensure the stability and consistency of the final light intensity output. By accurately controlling the light intensity in real time, it is not only possible to maintain the consistency of the lighting effect and meet the lighting needs of different scenes, but also to save energy consumption and improve the energy efficiency performance of the overall system. BRIEF DESCRIPTION OF THE DRAWINGS
[0056] Figure 1 A schematic diagram of an embodiment of a method for controlling light intensity of an LED module in an embodiment of the present invention;
[0057] Figure 2 Schematic diagram of an embodiment of the LED module light intensity control system in the embodiment of the present invention. DETAILED DESCRIPTION
[0058] Embodiments of the present invention provide a method and system for controlling the light intensity of an LED module. The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.
[0059] For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1 , an embodiment of the LED module light intensity control method in the embodiment of the present invention includes:
[0060] Step 101, obtaining initial light intensity data of an LED module to be controlled and initial state information of corresponding LED lamp beads, and classifying the initial light intensity data into light intensity types to obtain classified light intensity data; wherein the LED module to be controlled includes at least a plurality of LED lamp beads;
[0061] It is understandable that the execution subject of the present invention may be an LED module light intensity control system, or a terminal or a server, which is not limited here. The embodiment of the present invention is described by taking a server as the execution subject as an example.
[0062] Specifically, in the control system of the LED module, step 101 is to obtain and process the initial light intensity data and corresponding status information of each LED lamp bead in the module, so as to perform subsequent precise control work. The following are the specific technical details of this step:
[0063] Get the initial data of LED module
[0064] Light intensity data collection: Use a light intensity meter or similar sensing device to measure the current light intensity of each lamp bead in the LED module to be controlled. These devices are usually able to convert the received light signal into an electrical signal, which is then digitized and processed.
[0065] Status information recording: Synchronously record the initial status of each LED lamp bead, including on / off status, brightness level, color setting, etc. This information can be provided by the LED controller, or needs to be obtained from the LED module's drive circuit through an interface.
[0066] Data transmission: The collected light intensity data and status information are transmitted to the main control unit. Data collection software or microcontroller programming is required to implement this process, and data transmission is carried out through communication protocols (such as I2C, SPI, UART, etc.).
[0067] Light intensity data classification
[0068] Light intensity data preprocessing: In the main control unit, the received light intensity data is preliminarily cleaned and preprocessed. This includes data filtering, denoising, normalization and other operations to improve data quality.
[0069] Classification model implementation: Train or use a pre-defined classification model to classify light intensity into multiple categories. Classification can be based on light intensity range (for example, light intensity can be divided into weak light, medium light, and strong light), or other optical characteristics such as color temperature. The classification method can be a simple threshold judgment or a more complex machine learning algorithm, customized according to actual needs.
[0070] Classification data storage: Store the classified light intensity data and its classification results in a database or data structure for access and use in subsequent steps. Ensure that the data structure is easy to retrieve quickly to improve the efficiency of subsequent control command responses.
[0071] Step 102: Matching brightness adjustment features corresponding to the classified light intensity data in a preset database to obtain various brightness adjustment feature data;
[0072] Specifically, in step 102, the detailed implementation may include the following process:
[0073] Establish brightness feature database:
[0074] Create a brightness feature database, in which each record contains a set of detailed control parameters of LED lamp beads under a specific light intensity type, such as light intensity value, power consumption, voltage, current, color temperature, etc.
[0075] For example, for LED lamp beads in the "medium light" category, the database records that the voltage usually required at this light intensity is 3V and the current is 700mA.
[0076] Write database matching logic:
[0077] Design a query logic that can retrieve the corresponding brightness adjustment features in the database based on different light intensities and status information.
[0078] For example, if the intensity of an LED lamp is classified as "weak light", the query logic will search the database for all control parameter records marked as "weak light".
[0079] Extract brightness adjustment features:
[0080] The brightness adjustment features are extracted from the records returned by the database. If the light intensity classification result of a certain LED lamp bead is "weak light", the extracted features are several set current values, which are designed to increase the brightness of the lamp bead to an appropriate level.
[0081] Optimize the organization of feature data:
[0082] The extracted brightness adjustment features are screened and organized to ensure that they can effectively cover each lamp bead that needs to be controlled in the LED module. If the LED lamp beads in an area share the same light intensity type, the brightness adjustment feature data can be used together.
[0083] Provides a reading interface for brightness adjustment characteristics:
[0084] Ensure that other parts of the LED control system (such as control algorithms or chips) can access these characteristic data efficiently. For example, an API interface can be provided through which the control algorithm reads the brightness adjustment parameters of a specific lamp bead.
[0085] Practical applications of the brightness adjustment feature:
[0086] Confirm that the control system can use the characteristic data to adjust the actual brightness of the LED lamp beads. Assuming that a group of LED lamp beads are classified as "strong light", their current values need to be appropriately reduced to reduce the brightness. The control system will adjust the PWM signal according to the characteristic parameters provided by the database, so that the current of these LED lamp beads is reduced and the brightness is reduced to a predetermined level.
[0087] Step 103: Based on the preset light intensity type weight, perform adjustment feature multi-modal fusion on each brightness adjustment feature data to obtain a fusion result, and generate a light intensity control instruction corresponding to the initial light intensity data based on the fusion result;
[0088] Specifically, in step 103, the process of generating the light intensity control instruction by multi-modal fusion of the brightness adjustment feature data can be implemented by the following more detailed steps:
[0089] Establish a weighting system:
[0090] Define a weight value for each light intensity type. For example, considering the user's preference for different lighting scenarios, you can set a high weight (such as weight 2) for "medium light" in a relaxed reading environment, and a low weight (such as weight 0.5) for the less used background lighting "weak light".
[0091] Design fusion algorithm:
[0092] Develop an algorithm that can fuse the brightness adjustment feature data according to the weights of different light intensity types. The algorithm uses, for example, the weighted average method, using the mathematical formula: Fusion parameter = (Eigenvalue1 * Weight1 + Eigenvalue2 * Weight2 + ...) / (Weight1 + Weight2 + ...).
[0093] Perform multimodal data fusion:
[0094] For each set of brightness adjustment feature data matched in the database, the fusion algorithm is executed according to its light intensity type. For example, if the two light intensity type feature values of a certain LED lamp are 400mA and 800mA, and the corresponding weights are 0.5 and 2, respectively, the fused feature data is (400 * 0.5 + 800 * 2) / (0.5 + 2) = 700mA.
[0095] Generate light intensity control instructions:
[0096] The specific control instructions are calculated based on the fused data. Taking the fused current value as an example, the generated light intensity control instruction includes an instruction to set the current drive value of the LED to 700mA.
[0097] Format control instructions:
[0098] The calculated control instructions are formatted into signals or codes that can be understood by the driver circuit. If the LED driver uses digital signals to control the current, the current value of 700mA needs to be converted into the corresponding digital control code.
[0099] Transmission control instructions:
[0100] The formatted control instructions are sent to each LED driver through the communication interface of the LED control system, such as sending a command to set the current value to the driver through the I2C bus.
[0101] Real-time monitoring and feedback adjustment:
[0102] Implement real-time monitoring to measure whether the actual light intensity output matches the expected value, and adjust the algorithm through feedback to refine the control process. For example, use a light intensity sensor to measure the actual light intensity. If the actual value is lower than expected, the light intensity is adjusted by increasing the control current.
[0103] Step 104, obtaining the current working state information of the variable frequency power supply of the LED module to be controlled, and dynamically adjusting the light intensity control instruction based on the current working state information of the variable frequency power supply of the LED module to be controlled to obtain the adjusted light intensity control instruction;
[0104] Specifically, step 104 involves dynamically adjusting the light intensity control instruction considering the variable frequency power supply state of the LED module to match the power supply state and protect the stability of the LED. The specific implementation of this process includes the following steps:
[0105] Monitor the variable frequency power supply status:
[0106] Use sensors or power monitoring devices to collect real-time information about the current working status of the variable frequency power supply used by the LED module. The monitored information includes the output voltage, current, frequency, temperature, etc. of the power supply.
[0107] For example, a sensor monitors in real time that the output current of the variable frequency power supply is 500mA on average within a specific time.
[0108] Data transmission to the control unit:
[0109] Ensure that the power status information can be transmitted to the central control unit or processor of the LED control system via an appropriate communication protocol. Use standardized communication protocols such as SPI and I2C to send the monitored data to the main controller.
[0110] Analyze power supply operating conditions:
[0111] The control unit software analyzes the received variable frequency power supply operating status data to determine whether it is in the optimal operating range or whether there is a risk of exceeding normal operating conditions.
[0112] For example, if the temperature of the variable frequency power supply exceeds a predetermined safety value (e.g. 80°C), the control unit will receive this signal and take further analysis.
[0113] Instruction adjustment logic design:
[0114] Design an adjustment logic or algorithm that adjusts the existing light intensity control instructions based on the analyzed power status information. The purpose of the adjustment is to optimize the performance of the LED module while protecting the variable frequency power supply and LED module from damage.
[0115] If the temperature is too high, reduce the current to cool down, or adjust the power supply frequency to stabilize the working state of the LED.
[0116] Execute dynamic adjustment of light intensity control instructions:
[0117] Combining the above fusion result (step 103) with the current working state of the variable frequency power supply, the control unit dynamically adjusts the light intensity control instruction of the LED to adapt to the change of the power supply state.
[0118] If the variable frequency power supply current is too high, the control unit can reduce the light intensity control command transmitted to the LED, for example, when the original command current is 700mA, it is adjusted to 650mA.
[0119] Update control command output:
[0120] The adjusted light intensity control command is updated to the controller output. The adjusted command needs to be formatted and encoded again to obtain a signal suitable for the LED module and variable frequency power supply.
[0121] The updated instructions will be sent to the LED drive circuit to adjust the actual working state of the LED module.
[0122] Feedback and optimization control strategy:
[0123] After the adjusted instructions take effect, the working status of the LED and variable frequency power supply is continuously monitored, and the feedback information is used to optimize the adjustment strategy to maintain the best working conditions.
[0124] As feedback data from actual power supplies and LED modules accumulate, adjustment strategies can be continuously optimized through methods such as machine learning to achieve more efficient and intelligent lighting control.
[0125] Step 105: Based on the adjusted light intensity control instruction and the initial state information of the LED lamp beads, the initial state of the LED lamp beads is adjusted to generate a light intensity control adjustment result, and based on the light intensity control adjustment result, the light intensity control adjustment is performed on the LED module to be controlled.
[0126] Specifically, in step 105, a more detailed implementation method for adjusting the initial state of the LED lamp beads can be described as follows:
[0127] Integrated light intensity control instructions and LED status data:
[0128] First, the adjusted light intensity control command is integrated into the management system of the LED module. For example, if certain LED lamp beads are set to produce cool white light and operate at a specific current, this information will be processed together with the adjusted light intensity control command.
[0129] Calculate brightness adjustment requirements:
[0130] The control algorithm is used to calculate the adjustment required from the current state to the target brightness state. For example, the target current value of a certain LED lamp bead is 700mA, but the initial state is 500mA. The algorithm determines that it needs to be increased by 200mA.
[0131] Design a gradual adjustment strategy:
[0132] In order to avoid discomfort to users or damage to LED lamp beads caused by rapid brightness changes, a gradual adjustment strategy is designed. This strategy can mitigate the brightness changes by only slightly adjusting the PWM duty cycle each time or gradually increasing the current in multiple steps.
[0133] Formatting adjustment command:
[0134] Convert the calculated adjustment into a specific electronic signal or code, such as an adjustment of a PWM waveform, to match the hardware or communication protocol used.
[0135] Implement light intensity adjustment:
[0136] The control system issues modification instructions to the LED module, and the driver circuit adjusts the brightness of the light source according to these instructions. During this process, the control system can gradually adjust the duty cycle, such as increasing the duty cycle by 5% in each adjustment cycle until the target brightness is reached.
[0137] Monitor real-time light intensity feedback:
[0138] At this stage, the control system will synchronously obtain feedback information from the light intensity sensor. The light intensity changes captured by the sensor will be used to calibrate the control command to ensure the accuracy of brightness adjustment.
[0139] Subsequent adjustments and optimizations:
[0140] Perform necessary brightness adjustment optimization based on real-time feedback. If the brightness is found to be lower than the requirement in some areas, increase the current adjustment appropriately, such as adding 10mA until the expected brightness is reached.
[0141] Save reconciliation records:
[0142] Once the brightness adjustment is completed, the system will store detailed information of the adjusted LED module, including the adjusted brightness level, color temperature and power parameters for future reference and maintenance.
[0143] Another embodiment of the LED module light intensity control method in the embodiment of the present invention includes: obtaining the initial light intensity data of the LED module to be controlled and the corresponding initial state information of the LED lamp beads, and classifying the initial light intensity data into light intensity types to obtain the classified light intensity data, including:
[0144] Performing data cleaning on the initial light intensity data to obtain the initial light intensity data after data cleaning; wherein the initial light intensity data at least includes brightness data, color temperature data, and color saturation data;
[0145] Normalizing the initial light intensity data after data cleaning to obtain normalized initial light intensity data;
[0146] Performing statistical analysis on the normalized initial light intensity data to obtain distribution characteristics of the initial light intensity data; wherein the distribution characteristics of the initial light intensity data at least include mean, variance, skewness, kurtosis, maximum value, and minimum value;
[0147] Performing feature extraction on the normalized initial light intensity data to obtain first feature data and second feature data; wherein the first feature data at least includes average amplitude, standard deviation, and root mean square value; and the second feature data at least includes power spectrum density, frequency distribution, and harmonic components;
[0148] Performing cluster analysis on the distribution characteristics of the initial light intensity data, the first feature data, and the second feature data to obtain a plurality of initial light intensity features;
[0149] Smoothing the multiple initial light intensity features, eliminating short-term fluctuations in the multiple initial light intensity features, and obtaining smoothed light intensity feature data;
[0150] The smoothed light intensity feature data is classified into light intensity types using a preset light intensity type classification model to obtain classified light intensity data; wherein the preset light intensity type classification model is a machine learning model obtained through advance training.
[0151] Specifically, in the embodiment of the LED module light intensity control method, the specific implementation details of each step are as follows:
[0152] Data cleaning:
[0153] Implement automated algorithms to identify and process outliers in the initial light intensity data, such as removing data points that are not within a reasonable physical range through threshold judgment and outlier pruning methods.
[0154] Use interpolation techniques, such as linear interpolation or Lagrange interpolation, to fill in the missing data caused by data cleaning.
[0155] Normalization:
[0156] A normalization algorithm is applied to the data to adjust the dimensions of brightness, color temperature, and color saturation to ensure uniformity and improve the stability of data processing.
[0157] The low-pass filter is used to mitigate the data noise and further refine the result of the normalization process to ensure no data dimension difference.
[0158] Statistical characteristics analysis:
[0159] Statistical software packages are used to calculate the distribution characteristics of normalized data to provide morphological description and distribution summary for the data.
[0160] Use graphical tools such as histograms and box plots to visualize the distribution characteristics of normalized data to provide support for subsequent analysis.
[0161] Feature extraction:
[0162] Perform frequency domain analysis, extract the frequency components of the data through Fourier transform, and explore the intrinsic rhythm and periodicity of the light intensity data.
[0163] The numerical calculation method is used to calculate the characteristic indicators such as the average amplitude, standard deviation, and root mean square value of each measurement point in the data.
[0164] Cluster analysis:
[0165] Apply dimensionality reduction techniques such as Multidimensional Scaling (MDS) or Principal Component Analysis (PCA) to improve clustering results.
[0166] Cluster verification can be done by using methods such as the Silhouette Coefficient to determine the final number and quality of clusters.
[0167] Feature smoothing:
[0168] Apply digital signal processing methods to filter and eliminate short-term fluctuations, which is the denoising process in signal processing.
[0169] Use a smoothing operator with a specific time window, such as Gaussian smoothing, to smooth the light intensity features point by point or segment by segment.
[0170] Classification model application:
[0171] Select a corresponding machine learning model that has been trained and verified with historical data and experimental results to improve classification accuracy.
[0172] The model is called to be applied to the smoothed light intensity feature data for classification, and a clear classification of the light intensity type is obtained based on the probability distribution or category label output by the model.
[0173] In the embodiments of the present invention, the beneficial effects are as follows: first, outliers and noise data are removed by data cleaning, which improves the accuracy and reliability of the data and provides a high-quality data basis for subsequent processing. Secondly, the normalization process converts the data into the same dimension, reduces the error caused by the dimension difference, and ensures the consistency of data processing and the reliability of the results. The distribution characteristics such as mean, variance, skewness, and kurtosis are obtained through statistical analysis, which provides a rich data basis for feature extraction and cluster analysis. Feature extraction includes average amplitude, standard deviation, root mean square value, power spectrum density, frequency distribution, and harmonic components, etc., which describe different aspects of LED light intensity in multiple dimensions. Cluster analysis mines the intrinsic relationships and patterns of data, groups similar light intensity features, and makes light intensity control more refined and intelligent. Data smoothing eliminates short-term fluctuations, makes data more stable and continuous, and improves the accuracy and stability of classification. The light intensity feature data is classified using a trained machine learning model, which improves classification efficiency and accuracy and reduces manual intervention. The comprehensive application of data cleaning, normalization, feature extraction, cluster analysis, smoothing and machine learning classification has significantly improved the accuracy and efficiency of LED module light intensity control, adapted to different application scenarios, and improved the performance and service life of LED modules.
[0174] Another embodiment of the LED module light intensity control method in the embodiment of the present invention includes: the training process of the machine learning model includes:
[0175] Collecting LED module light intensity training data under multiple different environmental conditions; wherein the LED module light intensity training data at least includes brightness training data, color temperature training data, and color saturation training data;
[0176] Obtaining a preset machine learning model; wherein the preset machine learning model includes at least an input layer, a multiple feature extraction layer, a feature embedding layer, a fusion layer, a multi-layer fully connected network layer, and a multi-level classification layer, and the multi-level classification layer includes at least a primary classifier and a refined classifier;
[0177] Inputting the LED module light intensity training data into the input layer for preprocessing to obtain the preprocessed LED module light intensity training data;
[0178] The pre-processed LED module light intensity training data is transmitted to the multiple feature extraction layers layer by layer to obtain a multi-dimensional feature vector;
[0179] The extracted multi-dimensional feature vector is input into the feature embedding layer, and the multi-dimensional feature vector is embedded in high dimension to obtain multiple high-dimensional embedding features;
[0180] The high-dimensional embedded features are input into the fusion layer, and each high-dimensional embedded feature is fused to obtain a comprehensive light intensity feature vector;
[0181] The comprehensive light intensity feature vector is input into a multi-layer fully connected network layer, and the comprehensive light intensity feature vector is extracted and combined layer by layer to obtain the light intensity feature representation;
[0182] Input the light intensity feature representation into the multi-level classification layer, perform preliminary classification on the light intensity feature representation through the primary classifier in the multi-level classification layer to obtain a preliminary classification result, and perform refinement classification on the preliminary classification result through the refinement classifier in the multi-level classification layer to obtain a light intensity type label;
[0183] During the training process, combined with the preset actual light intensity type label, the loss function of the model is continuously iteratively optimized through the preset back propagation algorithm, and the model parameters are adjusted until the loss function converges to obtain the trained machine learning model as the light intensity type classification model; wherein the light intensity type classification model is used to classify the smoothed light intensity feature data into light intensity types to obtain classified light intensity data.
[0184] Specifically, in the embodiment of the present invention, the implementation details of each step of the LED module light intensity control method are further added, including:
[0185] Collect LED module light intensity training data under multiple different environmental conditions:
[0186] Experimental environment setup: Build a laboratory environment that simulates different temperature, humidity, and illumination conditions. For example, set different temperatures and humidity by adjusting the air conditioner and humidifier, and set different illumination conditions by using an adjustable light source.
[0187] Optical sensor configuration: Choose a high-precision optical sensor, such as a spectrophotometer or spectrometer, to ensure that the brightness, color temperature, and color saturation of the LED module can be accurately measured under different conditions.
[0188] Data recording and organization: Use a data acquisition system (such as LabVIEW or homemade data logging software) to record the data from the optical sensor in real time, store this data together with the corresponding environmental conditions in a database, and annotate the data to form a structured training data set.
[0189] Get the preset machine learning model:
[0190] Framework selection and installation: Select a suitable deep learning framework according to specific needs, such as TensorFlow, PyTorch, and install the required libraries and dependencies.
[0191] Model architecture definition: Define the model architecture through code (such as Python), including: the input layer is used to receive training data, the multiple feature extraction layers extract features through convolutional neural networks (CNN), the feature embedding layer maps features to high-dimensional space through fully connected layers, the fusion layer performs feature fusion, the multi-layer fully connected network layer is used for further feature extraction and combination, and the multi-level classification layer performs classification.
[0192] Classifier configuration: Set the structure of the primary classifier and the refined classifier, including the number of neurons in each layer, activation function (such as ReLU, Sigmoid), loss function (such as cross entropy loss), and optimizer (such as Adam, SGD).
[0193] Input the LED module light intensity training data into the input layer for preprocessing:
[0194] Data normalization: Normalize the training data to make the data range of different features consistent, usually scaling the data to the range of [0, 1] or [-1, 1].
[0195] Data cleaning: Detect and deal with outliers and noise in the data, for example, by identifying outliers through statistical methods (such as Z-score), processing through interpolation methods or removing outlier data points.
[0196] Data formatting: Convert the preprocessed data into the format required for model input. For example, convert the data into tensor form and divide it into training set, validation set, and test set.
[0197] The pre-processed LED module light intensity training data is sent to multiple feature extraction layers layer by layer:
[0198] Convolutional layer construction: Set up multiple convolutional layers, each of which includes convolution operations, activation functions (such as ReLU) and batch normalization to extract local features.
[0199] Pooling layer configuration: Add a pooling layer (such as maximum pooling or average pooling) after the convolution layer to reduce the size of the feature map while retaining important features, improving the computational efficiency and anti-overfitting ability of the model.
[0200] Input the extracted multidimensional feature vector into the feature embedding layer:
[0201] Fully connected layer design: Design multiple fully connected layers to map multi-dimensional feature vectors to high-dimensional space. Add activation functions (such as ReLU) and Dropout layers (to prevent overfitting) after each fully connected layer.
[0202] Application of high-dimensional embedding technology: Use linear transformation (such as matrix multiplication) or nonlinear transformation (such as kernel method) to map feature vectors to high-dimensional space to increase the expressive power of the model.
[0203] Input the high-dimensional embedding features into the fusion layer:
[0204] Feature fusion method: Select an appropriate feature fusion method, such as weighted summation, concatenation, or attention mechanism, to achieve effective feature fusion through weighting coefficients or attention weights.
[0205] Fusion layer design: Implement the specific logic of the fusion layer through code to ensure that each high-dimensional embedded feature can be fully integrated to improve the accuracy and completeness of feature representation.
[0206] Input the comprehensive light intensity feature vector into the multi-layer fully connected network layer:
[0207] Network layer construction: Build a multi-layer fully connected network, each layer includes a fully connected operation, an activation function, and a regularization layer (such as L2 regularization), extract and combine features layer by layer, and improve the nonlinear expression ability of the model.
[0208] Introduction of nonlinearity: Add activation functions (such as ReLU and Leaky ReLU) after each layer of the fully connected network to introduce nonlinearity and improve the model's ability to fit complex data.
[0209] Input the light intensity feature representation into the multi-level classification layer:
[0210] Primary classifier implementation: Implement the primary classifier, use the Softmax function to calculate the probability of each category, and perform preliminary classification of the light intensity feature representation.
[0211] Refinement classifier configuration: Configure the refinement classifier to refine the preliminary classification results through finer-grained classifiers (such as more fully connected layers or higher-dimensional output nodes) to obtain accurate light intensity type labels.
[0212] Optimize the model during training:
[0213] Loss function calculation: Use the cross entropy loss function to calculate the error between the model's predicted label and the actual label.
[0214] Application of back-propagation algorithm: The back-propagation algorithm is used to calculate the gradient of the loss function with respect to the model parameters, and the optimizer (such as Adam and SGD) is used to update the model parameters.
[0215] Iterative training and convergence judgment: Set the number of training rounds and early stopping strategy (such as stopping training when the validation set loss no longer decreases), and continuously iterate the training until the loss function converges to obtain the trained machine learning model.
[0216] Model Application:
[0217] Feature data smoothing: Smoothing the light intensity feature data to remove noise and fluctuations to obtain smoothed light intensity feature data.
[0218] Application of classification model: Input the smoothed light intensity feature data into the trained light intensity type classification model, classify it through the model, obtain the classified light intensity data, and realize precise control of the light intensity of the LED module.
[0219] System integration and deployment: Integrate the trained model into the LED module control system, deploy the model based on the actual application scenario, and realize real-time light intensity control and adjustment.
[0220] In the embodiment of the present invention, the beneficial effect is: the LED module light intensity control method proposed in the embodiment of the present invention realizes the refined control of light intensity characteristics by introducing a machine learning model for light intensity classification. The method collects LED module light intensity training data under multiple different environmental conditions, including multi-dimensional feature data such as brightness, color temperature, and color saturation, and uses a multi-level machine learning model for training. The model extracts and combines features layer by layer from the input layer to the multi-layer fully connected network layer, and finally realizes the preliminary classification and refined classification of light intensity characteristics through a multi-level classification layer. By iteratively optimizing the back propagation algorithm in combination with the actual light intensity type label, a converged light intensity type classification model is finally obtained. This embodiment: First, through multiple feature extraction layers and feature embedding layers, high-dimensional embedding and fusion processing of light intensity features are realized, so that the model can understand and represent the light intensity features more comprehensively; secondly, through the primary classifier and refinement classifier of the multi-level classification layer, the accuracy and refinement of light intensity classification are improved, which can better adapt to the complex and changeable actual use environment; thirdly, the preset back propagation algorithm is used to optimize the model loss function, so that the training process is more efficient, and the final classification model has higher reliability and accuracy; finally, the method realizes the accurate classification of smoothed light intensity feature data, improves the accuracy and stability of LED module light intensity control, and thus provides better lighting effects and user experience in practical applications.
[0221] Another embodiment of the LED module light intensity control method in the embodiment of the present invention includes: performing adjustment feature multimodal fusion on each brightness adjustment feature data based on a preset light intensity type weight to obtain a fusion result, and generating a light intensity control instruction corresponding to the initial light intensity data based on the fusion result, including:
[0222] Obtaining weight parameters of multiple light intensity types; wherein the weight parameters are set based on preset environmental requirements and light intensity adjustment targets, and stored in a database;
[0223] The weight parameters are used to assign weights to the brightness adjustment feature data, and the contribution of the feature values of the brightness adjustment feature data is adjusted according to the weights of different light intensity types to generate a weighted feature matrix;
[0224] The weighted feature matrix is fused to obtain a comprehensive adjustment feature vector as a fusion result; wherein the fusion process includes a data weighted average method and a modality cross fusion method;
[0225] The fusion result is input into a preset light intensity control instruction generation algorithm to generate a corresponding light intensity control instruction.
[0226] Specifically, based on the preset light intensity type weight, each brightness adjustment feature data is subjected to adjustment feature multimodal fusion to obtain a fusion result, and based on the fusion result, the specific steps of generating a light intensity control instruction corresponding to the initial light intensity data are as follows:
[0227] Get weight parameters of multiple light intensity types:
[0228] Weight parameter design: According to different environmental requirements and light intensity adjustment targets, design weight parameters for multiple light intensity types. For example, for a reading environment, design a higher brightness weight; for a leisure environment, design a lower brightness weight and a higher color temperature weight.
[0229] Parameter storage: Store the designed weight parameters in the database. The database can be SQL or NoSQL database. Select the appropriate storage solution based on system requirements.
[0230] Parameter reading: By writing a data reading interface (such as SQLAlchemy or Pandas reading function in Python), the weight parameters are read from the database and loaded into memory when the model is running.
[0231] Use the weight parameter to assign weights to the brightness adjustment feature data:
[0232] Feature data normalization: Normalize the brightness adjustment feature data and scale the data to a uniform range (such as [0,1]) to ensure that the weights of different features are allocated on the same scale.
[0233] Weight distribution calculation: According to the weights of different light intensity types, the eigenvalue of the brightness adjustment feature data is multiplied by the corresponding weight parameter to calculate the contribution of each eigenvalue. For example, for the brightness feature data X, the weight parameter is W, then the weighted eigenvalue is X * W.
[0234] Weighted feature matrix generation: The calculated weighted eigenvalues are combined into a weighted feature matrix for subsequent fusion processing.
[0235] Perform fusion processing on the weighted feature matrix:
[0236] Fusion method selection: Select a suitable fusion method, such as data weighted average method or modal cross fusion method. The weighted average method averages the weighted feature values; the modal cross fusion method cross-calculates the features of different modalities to improve the fusion effect.
[0237] Data weighted average method implementation: Perform weighted average calculation on each eigenvalue in the weighted feature matrix to obtain a comprehensive adjustment eigenvector. The specific implementation includes weighted summing each eigenvalue and then dividing it by the weighted sum.
[0238] Implementation of modal cross-fusion method: Modal cross-fusion is implemented by writing code, and the features of different modalities are cross-operated, such as multiplying and adding different eigenvalues to obtain a comprehensive adjusted feature vector.
[0239] Input the fusion result into the preset light intensity control instruction generation algorithm:
[0240] Preprocessing of fusion results: Preprocess the comprehensive adjustment feature vector, such as normalization or standardization, to ensure stability and consistency when the data is input into the model.
[0241] Light intensity control command generation algorithm design: Design a light intensity control command generation algorithm according to system requirements, such as using a linear regression model or a neural network model to generate light intensity control commands based on the fusion results.
[0242] Instruction generation and output: The preprocessed fusion results are input into the light intensity control instruction generation algorithm, the corresponding light intensity control instructions are generated through model calculation, and output to the LED module control system to achieve real-time adjustment and control of light intensity.
[0243] System Integration and Optimization:
[0244] Model deployment and integration: Deploy the trained light intensity control command generation algorithm into the LED module control system and integrate it into the actual application environment to ensure that the system can respond and adjust the light intensity in real time.
[0245] Real-time monitoring and feedback: By real-time monitoring of the light intensity regulation effect of the system, collecting feedback data from actual applications, and continuously optimizing and adjusting weight parameters and control algorithms, the accuracy and response speed of the system can be improved.
[0246] System upgrade and maintenance: Regularly upgrade and maintain the system, adjust and optimize the light intensity control strategy based on actual usage and user feedback to ensure system stability and efficiency.
[0247] In the embodiment of the present invention, the beneficial effect is: the LED module light intensity control method proposed in the embodiment of the present invention realizes the refined control and adjustment of the light intensity characteristics by introducing a machine learning model for light intensity classification and weight allocation. The method first collects the LED module light intensity training data under multiple different environmental conditions, including multi-dimensional feature data such as brightness, color temperature, and color saturation, and uses a multi-level machine learning model for training. The model extracts and combines features layer by layer from the input layer to the multi-layer fully connected network layer, and finally realizes the preliminary classification and refined classification of the light intensity characteristics through a multi-level classification layer. On this basis, based on the preset light intensity type weights, the brightness adjustment feature data is multimodally fused to generate a comprehensive adjustment feature vector, and a corresponding light intensity control instruction is generated. This embodiment: First, through multiple feature extraction layers and feature embedding layers, high-dimensional embedding and fusion processing of light intensity features are achieved, so that the model can understand and represent light intensity features more comprehensively; second, through the primary classifier and refinement classifier of the multi-level classification layer, the accuracy and refinement of light intensity classification are improved, which can better adapt to the complex and changeable actual use environment; third, the preset back propagation algorithm is used to optimize the model loss function, so that the training process is more efficient, and the final classification model has higher reliability and accuracy; in addition, by weighting and multi-modal fusion of brightness adjustment feature data, precise adjustment of light intensity control is achieved, and the accuracy and stability of LED module light intensity control are improved, which can better meet different environmental requirements and light intensity adjustment goals, thereby providing better lighting effects and user experience in practical applications. This method significantly improves the performance of LED modules and has high application value.
[0248] Another embodiment of the LED module light intensity control method in the embodiment of the present invention includes: after the step of inputting the fusion result into a preset light intensity control instruction generation algorithm to generate a corresponding light intensity control instruction, it includes:
[0249] According to the light intensity control instruction, the light intensity parameters of the LED module are controlled, and the initial light intensity data of the LED lamp beads are adjusted in real time;
[0250] In the real-time adjustment process, the current light intensity data of the LED module is monitored and collected based on the preset feedback loop, compared with the initial light intensity data, and the weight parameters and characteristic values of the brightness adjustment characteristic data are adjusted;
[0251] Based on the adjusted weight parameters and the adjusted characteristic values of the brightness adjustment characteristic data, the generated light intensity control instructions are iteratively adjusted and optimized to obtain the optimized light intensity control instructions, and the optimized light intensity control instructions are applied to the LED module.
[0252] Specifically, in the embodiment of the present invention, the implementation details of each step of the LED module light intensity control method are further added, including:
[0253] Input the fusion result into the preset light intensity control instruction generation algorithm:
[0254] Preprocessing of fusion results: Preprocessing of the comprehensive adjustment feature vector usually includes normalization or standardization to ensure the consistency of the input data.
[0255] Light intensity control command generation algorithm design: Design a light intensity control command generation algorithm based on system requirements, which can use linear regression model, decision tree model or neural network model, etc. The model generates light intensity control commands suitable for the current environment by learning the relationship between the fusion results and the actual light intensity control commands.
[0256] Model training: Use historical data to train the light intensity control instruction generation algorithm, optimize model parameters, and improve the model's prediction accuracy.
[0257] According to the light intensity control instruction, the light intensity parameters of the LED module are controlled, and the initial light intensity data of the LED lamp beads are adjusted in real time:
[0258] Instruction parsing and execution: The generated light intensity control instructions are parsed into specific control parameters, such as brightness, current, etc., and the light intensity of the LED module is adjusted through the LED driver.
[0259] Real-time adjustment mechanism: When executing the light intensity control instruction, the light intensity changes of the LED module are monitored in real time to ensure the stability and accuracy of the light intensity adjustment process.
[0260] In the real-time adjustment process, the current light intensity data of the LED module is monitored and collected based on the preset feedback loop, compared with the initial light intensity data, and the weight parameters and characteristic values of the brightness adjustment characteristic data are adjusted:
[0261] Feedback loop design: Establish a real-time feedback loop and use optical sensors to monitor the current light intensity data of the LED module in real time to ensure the continuity and accuracy of data collection.
[0262] Data comparison and analysis: Compare the current light intensity data with the initial light intensity data, analyze the light intensity change trend, and determine whether the weight parameters or eigenvalues need to be adjusted.
[0263] Weight parameter adjustment: According to the real-time data analysis results, dynamically adjust the weight parameters of the light intensity type to optimize the light intensity control effect.
[0264] Eigenvalue adjustment: Based on the adjusted weight parameters, the eigenvalues of the brightness adjustment feature data are recalculated to improve the accuracy of light intensity control.
[0265] Based on the adjusted weight parameters and the adjusted characteristic values of the brightness adjustment characteristic data, the generated light intensity control instructions are iteratively adjusted and optimized to obtain the optimized light intensity control instructions, and the optimized light intensity control instructions are applied to the LED module:
[0266] Iterative optimization process: Use iterative optimization algorithms (such as gradient descent method) to continuously adjust and optimize the light intensity control instructions based on real-time monitoring data to ensure the accuracy and stability of light intensity control.
[0267] Optimization model application: Apply the optimized light intensity control instructions to the LED module, adjust the light intensity parameters in real time, and ensure that the light intensity control effect meets expectations.
[0268] System verification and testing: Test the optimized light intensity control instructions under different environmental conditions to verify the stability and adaptability of the system and ensure the reliability of the system in various usage scenarios.
[0269] In the embodiment of the present invention, the beneficial effect is as follows: the LED module light intensity control method proposed in the embodiment of the present invention realizes the refined control and adjustment of the light intensity characteristics by introducing a machine learning model for light intensity classification and weight allocation. The method first collects the light intensity training data of the LED module under multiple different environmental conditions, including multi-dimensional feature data such as brightness, color temperature, and color saturation, and uses a multi-level machine learning model for training. From the input layer to the multi-layer fully connected network layer, the model extracts and combines features layer by layer, and finally realizes the preliminary classification and refined classification of the light intensity characteristics through a multi-level classification layer. On this basis, based on the preset light intensity type weights, the brightness adjustment feature data is multimodally fused to generate a comprehensive adjustment feature vector, and a corresponding light intensity control instruction is generated. During the real-time adjustment process, the current light intensity data is monitored and collected through a feedback loop, compared with the initial light intensity data, the weight parameters and feature values are adjusted, and the light intensity control instructions are iteratively optimized to ensure that the light intensity parameters of the LED module are adjusted to the optimal state in real time. This embodiment: First, through multiple feature extraction layers and feature embedding layers, high-dimensional embedding and fusion processing of light intensity features are achieved, so that the model can understand and represent light intensity features more comprehensively; secondly, through the primary classifier and refinement classifier of the multi-level classification layer, the accuracy and refinement of light intensity classification are improved, which can better adapt to the complex and changeable actual use environment; thirdly, the preset back propagation algorithm is used to optimize the model loss function, so that the training process is more efficient, and the final classification model has higher reliability and accuracy; in addition, by weighting and multi-modal fusion of brightness adjustment feature data, precise adjustment of light intensity control is achieved, and the accuracy and stability of light intensity control of LED modules are improved; finally, combined with real-time feedback and iterative optimization, the adaptability and response speed of the system are further enhanced, which can better meet the needs of different environments and light intensity adjustment goals, thereby providing better lighting effects and user experience in practical applications. This method significantly improves the performance of LED modules and has high application value.
[0270] Another embodiment of the LED module light intensity control method in the embodiment of the present invention includes: after the step of inputting the fusion result into a preset light intensity control instruction generation algorithm to generate a corresponding light intensity control instruction, it also includes:
[0271] Convert the generated light intensity control instruction into a character string in a preset format to obtain a character string of the light intensity control instruction;
[0272] Dynamically generate encryption keys according to preset encryption key generation rules; wherein the generated encryption keys are regularly updated based on preset security policies and environmental requirements;
[0273] The character string of the light intensity control instruction is encrypted using the generated encryption key to obtain the encrypted light intensity control instruction;
[0274] The encrypted light intensity control instruction is stored in a database, and a corresponding verification code is generated; wherein the verification code is used to ensure the data integrity and correctness of the encrypted light intensity control instruction during the transmission process.
[0275] Specifically, in the embodiment of the present invention, the implementation details of each step of the LED module light intensity control method are further added, including:
[0276] Input the fusion result into the preset light intensity control instruction generation algorithm:
[0277] Preprocessing of fusion results: Normalize the comprehensive adjustment feature vector to ensure the consistency of input data and the stability of model prediction.
[0278] Light intensity control command generation algorithm design: Select an appropriate machine learning model (such as linear regression, decision tree, or neural network) and train it so that it can generate accurate light intensity control commands based on the input fusion results.
[0279] Model training: Use historical data for training and optimize model parameters to ensure that the model can accurately generate light intensity control instructions.
[0280] According to the light intensity control instruction, the light intensity parameters of the LED module are controlled, and the initial light intensity data of the LED lamp beads are adjusted in real time:
[0281] Instruction parsing and execution: The generated light intensity control instructions are parsed into specific control parameters, such as brightness, color temperature, etc., and real-time control of the LED module is achieved through the LED drive circuit.
[0282] Real-time adjustment mechanism: During the execution process, the light intensity changes of the LED module are monitored in real time to ensure the stability and accuracy of the adjustment process.
[0283] In the real-time adjustment process, the current light intensity data of the LED module is monitored and collected based on the preset feedback loop, compared with the initial light intensity data, and the weight parameters and characteristic values of the brightness adjustment characteristic data are adjusted:
[0284] Feedback loop design: Use optical sensors to monitor the light intensity data of the LED module in real time and establish a feedback loop system.
[0285] Data comparison and analysis: Compare and analyze the monitored light intensity data with the initial data in real time to determine whether the current light intensity meets the expected target.
[0286] Weight parameter adjustment: adjust the weight parameters of the light intensity type according to the feedback data and dynamically optimize the control strategy.
[0287] Eigenvalue adjustment: Recalculate the eigenvalue of the brightness adjustment characteristic data to ensure the accuracy and stability of light intensity control.
[0288] Based on the adjusted weight parameters and the adjusted characteristic values of the brightness adjustment characteristic data, the generated light intensity control instructions are iteratively adjusted and optimized to obtain the optimized light intensity control instructions, and the optimized light intensity control instructions are applied to the LED module:
[0289] Iterative optimization process: Iterative optimization algorithms (such as gradient descent method) are used to continuously adjust and optimize the light intensity control instructions to ensure the accuracy of the control effect.
[0290] Optimization model application: Apply the optimized light intensity control instructions to the LED module to ensure that the actual effect of light intensity control is consistent with the expected target.
[0291] System verification and testing: Test the optimized light intensity control instructions under various environmental conditions to verify their stability and adaptability.
[0292] Convert the generated light intensity control instruction into a string in a preset format to obtain the string of the light intensity control instruction:
[0293] Instruction formatting: Convert the generated light intensity control instructions into a string according to the preset format to facilitate subsequent processing and transmission.
[0294] Dynamically generate encryption keys according to the preset encryption key generation rules:
[0295] Encryption key generation: Dynamically generate encryption keys based on preset security policies and environmental requirements. The generation rules can be updated regularly based on timestamps, random numbers or other security factors to ensure the security of the keys.
[0296] Key management: The generated encryption keys are securely stored in a key management system to ensure their availability and security.
[0297] The generated encryption key is used to encrypt the string of the light intensity control instruction to obtain the encrypted light intensity control instruction:
[0298] Encryption process: Use symmetric or asymmetric encryption algorithms (such as AES, RSA) to encrypt the string of light intensity control instructions to ensure the security of the instructions during transmission.
[0299] Encryption algorithm implementation: By writing the implementation code of the encryption algorithm, the efficiency and security of the encryption process can be ensured.
[0300] The encrypted light intensity control instruction is stored in the database, and the corresponding verification code is generated:
[0301] Instruction storage: Encrypted light intensity control instructions are stored in a secure database to ensure data persistence and availability.
[0302] Verification code generation: A verification code is generated through a hash algorithm (such as SHA-256) to verify the integrity and correctness of the instruction during transmission.
[0303] Verification and authentication: During the instruction transmission process, the integrity and correctness of the instruction are verified through the checksum to ensure that the instruction has not been tampered with.
[0304] In the embodiment of the present invention, the beneficial effect is as follows: the LED module light intensity control method proposed in the embodiment of the present invention realizes the refined control and adjustment of the light intensity feature by introducing a machine learning model for light intensity classification and weight allocation. The method first collects the light intensity training data of the LED module under multiple different environmental conditions, including multi-dimensional feature data such as brightness, color temperature, and color saturation, and uses a multi-level machine learning model for training. The model extracts and combines features layer by layer from the input layer to the multi-layer fully connected network layer, and finally realizes the preliminary classification and refined classification of the light intensity feature through the multi-level classification layer. On this basis, based on the preset light intensity type weight, the brightness adjustment feature data is multimodally fused to generate a comprehensive adjustment feature vector, and the corresponding light intensity control instruction is generated. In the real-time adjustment process, the current light intensity data is monitored and collected through the feedback loop, compared with the initial light intensity data, the weight parameters and feature values are adjusted, and the light intensity control instruction is iteratively optimized to ensure that the light intensity parameters of the LED module are adjusted to the optimal state in real time. Subsequently, the generated light intensity control instruction is converted into a string in a preset format and encrypted by a dynamically generated encryption key to ensure the integrity and security of data transmission. This embodiment: First, through multiple feature extraction layers and feature embedding layers, high-dimensional embedding and fusion processing of light intensity features are achieved, so that the model can understand and represent light intensity features more comprehensively; secondly, through the primary classifier and refinement classifier of the multi-level classification layer, the accuracy and refinement of light intensity classification are improved, which can better adapt to the complex and changeable actual use environment; thirdly, the preset back propagation algorithm is used to optimize the model loss function, so that the training process is more efficient, and the final classification model has higher reliability and accuracy; in addition, by weighting and multimodal fusion of brightness adjustment feature data, precise adjustment of light intensity control is achieved, and the accuracy and stability of light intensity control of LED modules are improved; finally, combined with real-time feedback and iterative optimization, the adaptability and response speed of the system are further enhanced, and the security and integrity of light intensity control instructions during transmission are ensured through encryption mechanism.
[0305] Another embodiment of the LED module light intensity control method in the embodiment of the present invention includes: after the step of storing the encrypted light intensity control instruction in the database and generating the corresponding verification code, it includes:
[0306] The transmission path is encrypted by a preset secure communication protocol to obtain an encrypted transmission path; wherein the encrypted transmission path is used to securely transmit the encrypted light intensity control instruction and the corresponding verification code to the LED module to prevent it from being obtained or tampered with by unauthorized parties;
[0307] After the LED module receives the encrypted light intensity control instruction and the corresponding verification code, the LED module is verified based on the corresponding verification code. If the verification passes, the LED module is authorized to start the decryption program; wherein the decryption program includes: using a preset decryption key generation rule to dynamically generate a decryption key, decrypting the encrypted light intensity control instruction based on the generated decryption key, and obtaining the decrypted light intensity control instruction.
[0308] Specifically, in the embodiment of the present invention, the implementation details of each step of the LED module light intensity control method are further added, including:
[0309] The transmission path is encrypted through the preset secure communication protocol to obtain the encrypted transmission path:
[0310] Secure communication protocol selection: Select an appropriate secure communication protocol (such as TLS / SSL) to ensure the security of data transmission.
[0311] Communication protocol configuration: Configure security certificates and keys on both communicating parties (server and LED module controller) to establish an encrypted communication channel.
[0312] Data encryption transmission: The encrypted light intensity control instructions and checksums are sent to the LED module through an encrypted transmission path to ensure the confidentiality and integrity of the data during transmission.
[0313] After the LED module receives the encrypted light intensity control instruction and the corresponding verification code, the LED module is verified based on the corresponding verification code:
[0314] Checksum verification: After the LED module receives the data, it first uses a hash algorithm (such as SHA-256) to calculate the checksum of the received data and compares it with the received checksum.
[0315] Verification pass judgment: If the verification passes, it means that the data has not been tampered with during the transmission process, and the LED module will authorize the start of the decryption program; if the verification fails, the data will be discarded and the abnormal situation will be recorded.
[0316] Authorize the LED module to start the decryption program:
[0317] Decryption key generation rules: Decryption keys are dynamically generated according to preset decryption key generation rules. The generation rules can be based on factors such as timestamps and random numbers to ensure the security of the keys.
[0318] Key management and storage: Ensure that the generated decryption key is securely stored in the LED module memory and destroyed promptly after use to prevent key leakage.
[0319] The encrypted light intensity control instruction is decrypted based on the generated decryption key to obtain the decrypted light intensity control instruction:
[0320] Decryption process implementation: Use a decryption algorithm that is symmetric to the encryption process (such as the AES decryption algorithm) and write decryption code to ensure the efficiency and correctness of the decryption process.
[0321] Instruction parsing and execution: The decrypted light intensity control instructions are parsed into specific control parameters, such as brightness, color temperature, etc., and real-time control of the LED module is achieved through the LED drive circuit.
[0322] In the embodiment of the present invention, the beneficial effect is as follows: the LED module light intensity control method proposed in the embodiment of the present invention realizes the refined control and adjustment of the light intensity feature by introducing a machine learning model for light intensity classification and weight allocation. The method first collects the light intensity training data of the LED module under multiple different environmental conditions, including multi-dimensional feature data such as brightness, color temperature, and color saturation, and uses a multi-level machine learning model for training. The model extracts and combines features layer by layer from the input layer to the multi-layer fully connected network layer, and finally realizes the preliminary classification and refined classification of the light intensity feature through the multi-level classification layer. On this basis, based on the preset light intensity type weight, the brightness adjustment feature data is multimodally fused to generate a comprehensive adjustment feature vector, and the corresponding light intensity control instruction is generated. In the real-time adjustment process, the current light intensity data is monitored and collected through the feedback loop, compared with the initial light intensity data, the weight parameters and feature values are adjusted, and the light intensity control instruction is iteratively optimized to ensure that the light intensity parameters of the LED module are adjusted to the optimal state in real time. Subsequently, the generated light intensity control instruction is converted into a string in a preset format and encrypted by a dynamically generated encryption key to ensure the integrity and security of data transmission. After the LED module receives the encrypted light intensity control instruction and check code, it ensures the correct interpretation and execution of the control instruction through the secure communication protocol and decryption program. This embodiment: First, through multiple feature extraction layers and feature embedding layers, high-dimensional embedding and fusion processing of light intensity features are realized, so that the model can understand and represent light intensity features more comprehensively; secondly, through the primary classifier and refinement classifier of the multi-level classification layer, the accuracy and refinement of light intensity classification are improved, which can better adapt to the complex and changeable actual use environment; thirdly, the preset back propagation algorithm is used to optimize the model loss function, so that the training process is more efficient, and the final classification model has higher reliability and accuracy; in addition, by weighting and multimodal fusion of brightness adjustment feature data, precise adjustment of light intensity control is achieved, and the accuracy and stability of light intensity control of LED modules are improved; finally, combined with real-time feedback and iterative optimization, the adaptability and response speed of the system are further enhanced, and the security and integrity of light intensity control instructions during transmission are ensured through encryption mechanism and secure communication protocol.
[0323] The above describes the LED module light intensity control method in the embodiment of the present invention. The following describes the LED module light intensity control system in the embodiment of the present invention. Figure 2 In one embodiment of the present invention, an LED module light intensity control system includes:
[0324] An acquisition module is used to acquire initial light intensity data of the LED module to be controlled and initial state information of the corresponding LED lamp beads, and classify the initial light intensity data into light intensity types to obtain classified light intensity data; wherein the LED module to be controlled includes at least a plurality of LED lamp beads;
[0325] A matching module, used to match the brightness adjustment features corresponding to the classified light intensity data in a preset database to obtain each brightness adjustment feature data;
[0326] A fusion module is used to perform multi-modal fusion of adjustment features on each brightness adjustment feature data based on a preset light intensity type weight to obtain a fusion result, and generate a light intensity control instruction corresponding to the initial light intensity data based on the fusion result;
[0327] An adjustment module, used for acquiring current working state information of the variable frequency power supply of the LED module to be controlled, and dynamically adjusting the light intensity control instruction based on the current working state information of the variable frequency power supply of the LED module to be controlled to obtain an adjusted light intensity control instruction;
[0328] The control and adjustment module is used to adjust the initial state of the LED lamp beads based on the adjusted light intensity control instructions and the initial state information of the LED lamp beads, generate a light intensity control adjustment result, and based on the light intensity control adjustment result, perform light intensity control adjustment on the LED module to be controlled.
[0329] The present invention also provides an LED module light intensity control device, which includes a memory and a processor. The memory stores computer-readable instructions. When the computer-readable instructions are executed by the processor, the processor executes the steps of the LED module light intensity control method in the above-mentioned embodiments.
[0330] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions are executed on a computer, the computer executes the steps of the LED module light intensity control method.
[0331] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.
[0332] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or the whole or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program code.
[0333] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A method for controlling light intensity of an LED module, characterized in that: The LED module light intensity control method comprises the following steps: Acquire the initial light intensity data of the LED module to be controlled and the corresponding initial state information of the LED lamp beads, and classify the initial light intensity data into light intensity types to obtain the classified light intensity data; wherein, the LED module to be controlled includes at least a plurality of LED lamp beads; the acquiring the initial light intensity data of the LED module to be controlled and the corresponding initial state information of the LED lamp beads, and classifying the initial light intensity data into light intensity types to obtain the classified light intensity data, comprises: performing data cleaning on the initial light intensity data to obtain the initial light intensity data after data cleaning; wherein, the initial light intensity data at least includes brightness data, color temperature data and color saturation data; performing normalization processing on the initial light intensity data after data cleaning to obtain the normalized initial light intensity data; performing statistical analysis on the normalized initial light intensity data to obtain the distribution characteristics of the initial light intensity data; wherein wherein the distribution characteristics of the initial light intensity data at least include mean, variance, skewness, kurtosis, maximum value and minimum value; feature extraction is performed on the normalized initial light intensity data to obtain first feature data and second feature data; wherein the first feature data at least includes average amplitude, standard deviation and root mean square value; the second feature data at least includes power spectrum density, frequency distribution and harmonic components; cluster analysis is performed on the distribution characteristics of the initial light intensity data, the first feature data and the second feature data to obtain a plurality of initial light intensity features; the plurality of initial light intensity features are smoothed to eliminate short-term fluctuations in the plurality of initial light intensity features to obtain smoothed light intensity feature data; the light intensity type classification is performed on the smoothed light intensity feature data using a preset light intensity type classification model to obtain classified light intensity data; wherein the preset light intensity type classification model is a machine learning model obtained through pre-training; The training process of the machine learning model includes: collecting LED module light intensity training data under multiple different environmental conditions; wherein the LED module light intensity training data at least includes brightness training data, color temperature training data and color saturation training data; obtaining a preset machine learning model; wherein the preset machine learning model at least includes an input layer, a multiple feature extraction layer, a feature embedding layer, a fusion layer, a multi-layer fully connected network layer and a multi-level classification layer, and the multi-level classification layer at least includes a primary classifier and a refined classifier; inputting the LED module light intensity training data into the input layer for preprocessing to obtain preprocessed LED module light intensity training data; transmitting the preprocessed LED module light intensity training data to the multiple feature extraction layers layer by layer to obtain a multi-dimensional feature vector; The extracted multidimensional feature vector is input into the feature embedding layer, and the multidimensional feature vector is embedded in high dimension to obtain multiple high-dimensional embedding features; the high-dimensional embedding features are input into the fusion layer, and each high-dimensional embedding feature is fused to obtain a comprehensive light intensity feature vector; the comprehensive light intensity feature vector is input into the multi-layer fully connected network layer, and the comprehensive light intensity feature vector is extracted and combined layer by layer to obtain the light intensity feature representation; the light intensity feature representation is input into the multi-level classification layer, and the light intensity feature representation is preliminarily classified by the primary classifier in the multi-level classification layer to obtain the preliminary classification result, and the preliminary classification result is refined by the refinement classifier in the multi-level classification layer to obtain the light intensity type label; in the training process, combined with the preset actual light intensity type label, the loss function of the model is continuously iteratively optimized by the preset back propagation algorithm, and the model parameters are adjusted until the loss function converges to obtain the trained machine learning model as the light intensity type classification model; wherein, the light intensity type classification model is used to classify the light intensity type of the smoothed light intensity feature data to obtain the classified light intensity data; Matching brightness adjustment features corresponding to the classified light intensity data in a preset database to obtain respective brightness adjustment feature data; Based on the preset light intensity type weight, each brightness adjustment feature data is subjected to adjustment feature multi-modal fusion to obtain a fusion result, and based on the fusion result, a light intensity control instruction corresponding to the initial light intensity data is generated; Acquire current working state information of the variable frequency power supply of the LED module to be controlled, and dynamically adjust the light intensity control instruction based on the current working state information of the variable frequency power supply of the LED module to be controlled to obtain an adjusted light intensity control instruction; Based on the adjusted light intensity control instruction and the initial state information of the LED lamp beads, the initial state of the LED lamp beads is adjusted to generate a light intensity control adjustment result, and based on the light intensity control adjustment result, the light intensity control adjustment is performed on the LED module to be controlled.
2. The LED module light intensity control method according to claim 1, characterized in that: The method of performing multi-modal fusion of adjustment features on each brightness adjustment feature data based on a preset light intensity type weight to obtain a fusion result, and generating a light intensity control instruction corresponding to the initial light intensity data based on the fusion result, includes: Obtaining weight parameters of multiple light intensity types; wherein the weight parameters are set based on preset environmental requirements and light intensity adjustment targets, and stored in a database; The weight parameters are used to assign weights to the brightness adjustment feature data, and the contribution of the feature values of the brightness adjustment feature data is adjusted according to the weights of different light intensity types to generate a weighted feature matrix; The weighted feature matrix is fused to obtain a comprehensive adjustment feature vector as a fusion result; wherein the fusion process includes a data weighted average method or a modality cross fusion method; The fusion result is input into a preset light intensity control instruction generation algorithm to generate a corresponding light intensity control instruction.
3. The LED module light intensity control method according to claim 2, characterized in that: After the step of inputting the fusion result into a preset light intensity control instruction generation algorithm to generate a corresponding light intensity control instruction, the method further comprises: According to the light intensity control instruction, the light intensity parameters of the LED module are controlled, and the initial light intensity data of the LED lamp beads are adjusted in real time; In the real-time adjustment process, the current light intensity data of the LED module is monitored and collected based on the preset feedback loop, compared with the initial light intensity data, and the weight parameters and characteristic values of the brightness adjustment characteristic data are adjusted; Based on the adjusted weight parameters and the adjusted characteristic values of the brightness adjustment characteristic data, the generated light intensity control instructions are iteratively adjusted and optimized to obtain the optimized light intensity control instructions, and the optimized light intensity control instructions are applied to the LED module.
4. The LED module light intensity control method according to claim 2, characterized in that: After the step of inputting the fusion result into a preset light intensity control instruction generation algorithm to generate a corresponding light intensity control instruction, the method further includes: Convert the generated light intensity control instruction into a character string in a preset format to obtain a character string of the light intensity control instruction; Dynamically generate encryption keys according to preset encryption key generation rules; wherein the generated encryption keys are regularly updated based on preset security policies and environmental requirements; The character string of the light intensity control instruction is encrypted using the generated encryption key to obtain the encrypted light intensity control instruction; The encrypted light intensity control instruction is stored in a database, and a corresponding verification code is generated; wherein the verification code is used to ensure the data integrity and correctness of the encrypted light intensity control instruction during the transmission process.
5. The LED module light intensity control method according to claim 4, characterized in that: After the step of storing the encrypted light intensity control instruction in the database and generating the corresponding verification code, the method further comprises: The transmission path is encrypted by a preset secure communication protocol to obtain an encrypted transmission path; wherein the encrypted transmission path is used to securely transmit the encrypted light intensity control instruction and the corresponding verification code to the LED module to prevent it from being obtained or tampered with by unauthorized parties; After the LED module receives the encrypted light intensity control instruction and the corresponding verification code, the LED module is verified based on the corresponding verification code. If the verification passes, the LED module is authorized to start the decryption program; wherein, the decryption program includes: using a preset decryption key generation rule to dynamically generate a decryption key, decrypting the encrypted light intensity control instruction based on the generated decryption key, and obtaining the decrypted light intensity control instruction.
6. A LED module light intensity control system, characterized in that: The LED module light intensity control system comprises: An acquisition module is used to acquire initial light intensity data of an LED module to be controlled and initial state information of corresponding LED lamp beads, and classify the initial light intensity data into light intensity types to obtain classified light intensity data; wherein, the LED module to be controlled includes at least a plurality of LED lamp beads; the acquisition of initial light intensity data of an LED module to be controlled and initial state information of corresponding LED lamp beads, and classify the initial light intensity data into light intensity types to obtain classified light intensity data, including: performing data cleaning on the initial light intensity data to obtain initial light intensity data after data cleaning; wherein, the initial light intensity data includes at least brightness data, color temperature data and color saturation data; performing normalization processing on the initial light intensity data after data cleaning to obtain normalized initial light intensity data; performing statistical analysis on the normalized initial light intensity data to obtain distribution characteristics of the initial light intensity data property; wherein the distribution characteristics of the initial light intensity data at least include mean, variance, skewness, kurtosis, maximum value and minimum value; feature extraction is performed on the normalized initial light intensity data to obtain first feature data and second feature data; wherein the first feature data at least includes average amplitude, standard deviation and root mean square value; the second feature data at least includes power spectrum density, frequency distribution and harmonic components; cluster analysis is performed on the distribution characteristics of the initial light intensity data, the first feature data and the second feature data to obtain multiple initial light intensity features; multiple initial light intensity features are smoothed to eliminate short-term fluctuations in multiple initial light intensity features to obtain smoothed light intensity feature data; light intensity type classification is performed on the smoothed light intensity feature data using a preset light intensity type classification model to obtain classified light intensity data; wherein the preset light intensity type classification model is a machine learning model obtained through pre-training; The training process of the machine learning model includes: collecting LED module light intensity training data under multiple different environmental conditions; wherein the LED module light intensity training data at least includes brightness training data, color temperature training data and color saturation training data; obtaining a preset machine learning model; wherein the preset machine learning model at least includes an input layer, a multiple feature extraction layer, a feature embedding layer, a fusion layer, a multi-layer fully connected network layer and a multi-level classification layer, and the multi-level classification layer at least includes a primary classifier and a refined classifier; inputting the LED module light intensity training data into the input layer for preprocessing to obtain preprocessed LED module light intensity training data; transmitting the preprocessed LED module light intensity training data to the multiple feature extraction layers layer by layer to obtain a multi-dimensional feature vector; The extracted multidimensional feature vector is input into the feature embedding layer, and the multidimensional feature vector is embedded in high dimension to obtain multiple high-dimensional embedding features; the high-dimensional embedding features are input into the fusion layer, and each high-dimensional embedding feature is fused to obtain a comprehensive light intensity feature vector; the comprehensive light intensity feature vector is input into the multi-layer fully connected network layer, and the comprehensive light intensity feature vector is extracted and combined layer by layer to obtain the light intensity feature representation; the light intensity feature representation is input into the multi-level classification layer, and the light intensity feature representation is preliminarily classified by the primary classifier in the multi-level classification layer to obtain the preliminary classification result, and the preliminary classification result is refined by the refinement classifier in the multi-level classification layer to obtain the light intensity type label; in the training process, combined with the preset actual light intensity type label, the loss function of the model is continuously iteratively optimized by the preset back propagation algorithm, and the model parameters are adjusted until the loss function converges to obtain the trained machine learning model as the light intensity type classification model; wherein, the light intensity type classification model is used to classify the light intensity type of the smoothed light intensity feature data to obtain the classified light intensity data; A matching module, used to match the brightness adjustment features corresponding to the classified light intensity data in a preset database to obtain various brightness adjustment feature data; A fusion module is used to perform multi-modal fusion of adjustment features on each brightness adjustment feature data based on a preset light intensity type weight to obtain a fusion result, and generate a light intensity control instruction corresponding to the initial light intensity data based on the fusion result; An adjustment module, used for acquiring current working state information of the variable frequency power supply of the LED module to be controlled, and dynamically adjusting the light intensity control instruction based on the current working state information of the variable frequency power supply of the LED module to be controlled to obtain an adjusted light intensity control instruction; The control and adjustment module is used to adjust the initial state of the LED lamp beads based on the adjusted light intensity control instructions and the initial state information of the LED lamp beads, generate a light intensity control adjustment result, and based on the light intensity control adjustment result, perform light intensity control adjustment on the LED module to be controlled.
7. A light intensity control device for an LED module, characterized in that: The LED module light intensity control device comprises: a memory and at least one processor, wherein instructions are stored in the memory; The at least one processor calls the instructions in the memory to enable the LED module light intensity control device to execute the LED module light intensity control method according to any one of claims 1-5.
8. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by the processor, the LED module light intensity control method according to any one of claims 1 to 5 is implemented.
