Laser galvanometer intelligent control system and method based on big data

Abstract

The invention provides a laser galvanometer intelligent control system and method based on big data, and belongs to the technical field of intelligent control systems. S2, data preprocessing; s3, feature extraction; s4, model training; s5, performing real-time prediction; s6, generating a control instruction; s7, executing a control instruction; and S8, performing feedback and optimization. High-precision closed-loop control over movement of the laser galvanometer can be achieved, the deviation between the actual movement state and the target movement state is effectively reduced, the laser machining precision is improved, and application scenes such as electronic chip fine machining with extremely high precision requirements are met.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent control systems, and more particularly to a laser galvanometer intelligent control system and method based on big data. Background Art

[0002] In modern industrial production, laser processing technology is widely used in many fields such as 3D printing, electronic manufacturing, and automotive parts processing due to its advantages of high precision, high efficiency, and non-contact. As a key execution component in the laser processing system, the motion control accuracy and speed of the laser galvanometer directly affect the quality and efficiency of laser processing. Traditional laser galvanometer control methods are mostly based on empirical models or simple control algorithms, which are difficult to cope with complex and changing working environments and the increasing demand for high-precision processing. Under complex working conditions, the laser galvanometer is easily interfered by environmental factors such as temperature and humidity, and as the complexity of the processing task increases, traditional methods are unable to adjust the motion state of the galvanometer in real time to ensure processing accuracy.

[0003] With the rapid development of emerging technologies such as big data and artificial intelligence, it provides a new opportunity for the innovation of laser galvanometer control technology. By introducing big data technology, a large amount of data generated during the operation of the laser galvanometer can be collected, analyzed, and processed, and the hidden laws behind the data can be mined, so as to achieve accurate prediction and intelligent control of the motion state of the laser galvanometer, make up for the deficiencies of traditional control methods, and meet the requirements of industrial production for higher quality and higher efficiency of laser processing. Summary of the Invention

[0004] The present invention provides a laser galvanometer intelligent control system and method based on big data to achieve high-precision intelligent control.

[0005] To achieve the above object, the present invention adopts the following technical solutions: A laser galvanometer intelligent control method based on big data, comprising: S1. Data acquisition; collecting the real-time position data, motion speed data, acceleration data, and environmental data of the laser galvanometer, respectively performing data packaging and timestamp marking processing to obtain position raw data, speed raw data, acceleration raw data, and environmental raw data; S2. Data preprocessing; deleting abnormal data and performing normalization processing on the position raw data, speed raw data, acceleration raw data, and environmental raw data to obtain position valid data, speed valid data, acceleration valid data, and environmental valid data, and integrating them to obtain a preprocessed data set; S3. Feature extraction: Extract the characteristics of the changing trend of the galvanometer position, the feature of the speed change rate, the feature of the acceleration change trend, and the relevant features of the influence of environmental factors on the performance of the galvanometer from the preprocessed dataset to form a feature vector set; Use the pendulum wave technique to process the feature vector set to obtain an enhanced feature vector set; S4. Model training: Divide the enhanced feature vector set into a training set and a test set; Use the training set to train the long short-term memory network model and the random forest model respectively; Weightedly fuse the trained long short-term memory network model and the random forest model to obtain a galvanometer motion prediction model; S5. Real-time prediction: Input the real-time enhanced feature vector set into the galvanometer motion prediction model to obtain the predicted position coordinates, speed, and acceleration of the galvanometer; S6. Control instruction generation: Calculate the position deviation, speed deviation, and acceleration deviation between the predicted position coordinates, speed, and acceleration of the galvanometer and the preset target motion state; Adopt the proportional-integral-differential control algorithm to calculate the control instructions for position, speed, and acceleration according to the above deviations to form a control instruction set; S7. Control instruction execution: Transmit the control instruction set to the control system of the galvanometer, adjust the motor drive parameters of the galvanometer, and achieve real-time control of the motion state of the galvanometer; S8. Feedback and optimization: After the galvanometer executes the control instruction, collect the actual motion state data again, and repeat the above steps S1-S7; Compare and analyze the newly collected data with the predicted data and the control target data. If the deviation between the actual motion state and the target motion state exceeds the preset error range, readjust the parameters in the proportional-integral-differential control algorithm and the parameters of the galvanometer motion prediction model to optimize the model.

[0006] In this specification, in S1, a fixed sampling frequency is set to periodically collect the position data, speed data, acceleration data, and environmental data of the galvanometer.

[0007] In this specification, in S2, the basis for judging abnormal data is that the position coordinates, speed, acceleration, or environmental data of the galvanometer exceed the preset reasonable range.

[0008] In this specification, in S3, the position change trend feature is obtained by calculating the difference in position coordinates at adjacent time points; the speed change rate feature is obtained by calculating the ratio of the difference in speed at adjacent time points to the time interval; the acceleration change trend feature is obtained by calculating the difference in acceleration at adjacent time points.

[0009] In this specification, in S3, the single pendulum wave technology simulates the periodic motion law of a simple pendulum to perform fluctuation analysis and enhancement on the data in the feature vector set, thereby mining the potential periodic characteristics of the data.

[0010] In this specification, in S4, the mean square error is used as the loss function during the training of the long short-term memory network model, and the model parameters are adjusted through the back propagation algorithm.

[0011] In this specification, in S4, the random forest model is composed of multiple decision trees. When constructing each decision tree, training samples are selected by random sampling with replacement, and some features are randomly selected for node splitting.

[0012] In this specification, in S4, when the long short-term memory network model and the random forest model are weightedly fused, the weighting coefficient is determined by adjusting the parameters on the validation set.

[0013] In this specification, in S8, a particle swarm optimization algorithm or a genetic algorithm is used to adjust the parameters in the proportional-integral-differential control algorithm and the parameters of the laser galvanometer motion prediction model.

[0014] A laser galvanometer intelligent control system based on big data, using any of the above-mentioned laser galvanometer intelligent control methods based on big data, the laser galvanometer intelligent control system based on big data comprising: The data acquisition module is used to collect the real-time position data, motion speed data, acceleration data, and environmental data of the laser galvanometer, and perform data packaging and time stamping processing respectively to obtain the original position data, original speed data, original acceleration data, and original environmental data; The data preprocessing module is used to delete abnormal data and normalize the original position data, original velocity data, original acceleration data and original environment data to obtain valid position data, valid velocity data, valid acceleration data and valid environment data, and integrate them to obtain the preprocessed data set; The feature extraction module is used to extract the laser galvanometer position change trend characteristics, velocity change rate characteristics, acceleration change trend characteristics, and related characteristics of the impact of environmental factors on the laser galvanometer performance from the preprocessed data set to form a feature vector set; the feature vector set is processed using the single pendulum wave technology to obtain an enhanced feature vector set; The model training module is used to divide the enhanced feature vector set into a training set and a test set; use the training set to train the long short-term memory network model and the random forest model respectively; and perform a weighted fusion of the trained long short-term memory network model and the random forest model to obtain a laser galvanometer motion prediction model; A real-time prediction module, which is used to input the real-time enhanced feature vector set into the galvanometer movement prediction model to obtain the predicted position coordinates, speed, and acceleration of the galvanometer; A control instruction generation module, which is used to calculate the position deviation, speed deviation, and acceleration deviation between the predicted position coordinates, speed, and acceleration of the galvanometer and the preset target motion state; and adopt a proportional-integral-derivative control algorithm to calculate the control instructions for position, speed, and acceleration according to the above deviations, and form a control instruction set; A control instruction execution module, which is used to transmit the control instruction set to the control system of the galvanometer, adjust the motor drive parameters of the galvanometer, and realize the real-time control of the motion state of the galvanometer; A feedback and optimization module, which is used to collect the actual motion state data of the galvanometer again after the galvanometer executes the control instruction, compare and analyze the newly collected data with the predicted data and the control target data. If the deviation between the actual motion state and the target motion state exceeds the preset error range, re-adjust the parameters in the proportional-integral-derivative control algorithm and the parameters of the galvanometer movement prediction model to optimize the model.

[0015] In summary, the present invention has at least the following beneficial effects: High-precision control: Through the collection, preprocessing, and in-depth analysis of multi-source data such as the position, speed, acceleration, and environmental data of the galvanometer, combined with advanced feature extraction methods and integrated learning models, the future motion state of the galvanometer can be accurately predicted. After comparing with the preset target motion state and using the PID control algorithm to generate control instructions, high-precision closed-loop control of the galvanometer motion can be realized, effectively reducing the deviation between the actual motion state and the target motion state, improving the accuracy of laser processing, and meeting application scenarios with extremely high precision requirements such as fine processing of electronic chips. Efficient real-time response: The entire control process from data collection to control instruction execution adopts efficient algorithms and process designs, can quickly process real-time data, and timely adjust the motion state of the galvanometer. In the face of dynamic changes in processing tasks, the system can quickly respond. For example, in the process of 3D printing, the model structure is complex and changeable, and this intelligent control method can quickly adjust the galvanometer motion to ensure the accuracy of the laser scanning path, greatly improving the processing efficiency. Stability improvement: The combination of the integrated LSTM model and the random forest model enhances the model's learning ability and generalization ability for complex data, making the prediction results more stable and reliable. At the same time, by regularly optimizing the model parameters and PID control parameters, as well as evaluating and optimizing the performance of decision trees in the random forest, it can effectively cope with the impact of factors such as changes in the working environment and equipment aging on the stability of galvanometer motion control, ensure the continuous and stable operation of the laser processing process, and reduce the defective rate of processing caused by equipment instability. Enhanced adaptability: The system has a feedback and optimization mechanism that can automatically adjust the control parameters according to the deviation between the actual motion state and the target state of the galvanometer, and has good adaptability to different working environments and processing tasks. Whether in harsh environments such as high temperature and high humidity, or facing processing objects of different materials and shapes, it can ensure the motion control effect of the galvanometer by continuously optimizing the control strategy, expanding the application range of laser processing technology. Description of the Drawings

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

[0017] Figure 1 It is a schematic flowchart of the intelligent control method for galvanometers based on big data involved in the present invention.

[0018] Figure 2 It is a schematic flowchart of feature extraction and enhancement involved in the present invention.

[0019] Figure 3 It is a schematic flowchart of model training and prediction involved in the present invention.

[0020] Figure 4 It is a schematic diagram of the intelligent control system for galvanometers based on big data involved in the present invention. Detailed Embodiments

[0021] In the following text, only some exemplary embodiments are simply described. As those skilled in the art can recognize, the described embodiments can be modified in various different ways without departing from the spirit or scope of the embodiments of the present invention. Therefore, the drawings and descriptions are considered to be exemplary in nature rather than restrictive.

[0022] The following disclosure provides many different embodiments or examples for implementing different structures of the embodiments of the present invention. To simplify the disclosure of the embodiments of the present invention, the components and settings of specific examples are described below. Of course, they are only examples and are not intended to limit the embodiments of the present invention. In addition, the embodiments of the present invention may repeat reference numerals and / or reference letters in different examples. This repetition is for the purpose of simplification and clarity, and does not itself indicate the relationship between the various embodiments and / or settings discussed.

[0023] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0024] As Figure 1 shown, this embodiment provides a laser galvanometer intelligent control method based on big data, including: S1. Data acquisition; collect the real-time position data, motion speed data, acceleration data, and environmental data of the laser galvanometer, and perform data packaging and timestamp marking processing respectively to obtain position raw data, speed raw data, acceleration raw data, and environmental raw data; S2. Data preprocessing; delete abnormal data and perform normalization processing on the position raw data, speed raw data, acceleration raw data, and environmental raw data to obtain position valid data, speed valid data, acceleration valid data, and environmental valid data, and integrate them to obtain a preprocessed dataset; S3. Feature extraction; extract the laser galvanometer position change trend features, speed change rate features, acceleration change trend features, and relevant features of the influence of environmental factors on the performance of the laser galvanometer from the preprocessed dataset to form a feature vector set; use the single pendulum wave technology to process the feature vector set to obtain an enhanced feature vector set; S4. Model training; divide the enhanced feature vector set into a training set and a test set; use the training set to train the long short-term memory network model and the random forest model respectively; perform weighted fusion on the trained long short-term memory network model and the random forest model to obtain a laser galvanometer motion prediction model; S5. Real-time prediction; input the real-time enhanced feature vector set into the laser galvanometer motion prediction model to obtain the predicted position coordinates, speed, and acceleration of the laser galvanometer; S6. Control instruction generation; calculate the position deviation, speed deviation, and acceleration deviation between the predicted position coordinates, speed, and acceleration of the laser galvanometer and the preset target motion state; adopt the proportional-integral-derivative control algorithm to calculate the control instructions for position, speed, and acceleration according to the above deviations to form a control instruction set; S7. Control instruction execution: transmitting the control instruction set to the laser galvanometer control system, adjusting the laser galvanometer motor drive parameters, and achieving real-time control of the laser galvanometer motion state; S8. Feedback and optimization; after the laser galvanometer executes the control instruction, its actual motion state data is collected again, and the above steps S1-S7 are repeated; the newly collected data is compared and analyzed with the predicted data and the control target data. If the deviation between the actual motion state and the target motion state exceeds the preset error range, the parameters in the proportional-integral-differential control algorithm and the parameters of the laser galvanometer motion prediction model are readjusted to optimize the model.

[0025] In some embodiments, in S1, a fixed sampling frequency is set to periodically collect position data, velocity data, acceleration data, and environmental data of the laser galvanometer.

[0026] In some embodiments, in S2, the abnormal data is determined based on whether the position coordinates, velocity, acceleration or environmental data of the laser galvanometer exceed a preset reasonable range.

[0027] In some embodiments, in S3, the position change trend feature is obtained by calculating the difference between the position coordinates of adjacent time points; the speed change rate feature is obtained by calculating the ratio of the difference between the speeds of adjacent time points to the time interval; and the acceleration change trend feature is obtained by calculating the difference between the accelerations of adjacent time points.

[0028] In some embodiments, in S3, the single pendulum wave technology simulates the periodic motion of a simple pendulum to perform fluctuation analysis and enhancement on the data in the feature vector set, thereby mining the potential periodic characteristics of the data.

[0029] In some embodiments, in S4, mean square error is used as a loss function during the training of the long short-term memory network model, and the model parameters are adjusted through a back propagation algorithm.

[0030] In some embodiments, in S4, the random forest model is composed of multiple decision trees, and each decision tree is constructed by using random sampling with replacement to select training samples and randomly selecting some features for node splitting.

[0031] In some embodiments, in S4, when the long short-term memory network model and the random forest model are weightedly fused, the weighting coefficient is determined by adjusting parameters on the validation set.

[0032] In some embodiments, in S8, a particle swarm optimization algorithm or a genetic algorithm is used to adjust the parameters in the proportional-integral-differential control algorithm and the parameters of the laser galvanometer motion prediction model.

[0033] In some embodiments, data acquisition: In the working environment of a galvanometer scanner, multiple sensors are used for data acquisition. Among them, a position sensor is used to collect the real-time position data of the galvanometer scanner, a speed sensor collects the motion speed data of the galvanometer scanner, an acceleration sensor collects the acceleration data of the galvanometer scanner, and an environment sensor collects environmental data such as temperature and humidity in the working environment. A fixed sampling frequency (unit: Hertz, Hz) is set, and the above data is collected periodically. Each set of collected data is marked with the acquisition time (unit: second, s) to form an original data set: ; where represents the three-dimensional position coordinates of the galvanometer scanner at the -th acquisition, represents the motion speed of the galvanometer scanner at the -th acquisition, represents the acceleration of the galvanometer scanner at the -th acquisition, represents the environmental data at the -th acquisition, represents the time of the -th acquisition, represents the total number of data acquisitions.

[0034] The process of data acquisition is as follows: According to the set sampling frequency , the position sensor, speed sensor, acceleration sensor, and environment sensor periodically obtain the corresponding data in chronological order, pack them in a fixed data format, add the acquisition timestamp and then store them in a temporary data storage area to form an original data set .

[0035] In some embodiments, data preprocessing: The original data set is transmitted to the data preprocessing module. First, data cleaning is performed to remove the abnormal data therein. The basis for judging abnormal data is: If any dimension value in the position coordinates of the galvanometer scanner exceeds the preset reasonable range , the speed

[0036] exceeds the range , the acceleration exceeds the range , or the environmental data If it exceeds the normal environmental parameter range (for example, the temperature or humidity exceeds the preset reasonable range), then the set of data is determined to be abnormal data and deleted. Then, the cleaned data is normalized. For position coordinate data, the formula is used: ; , , are the data after three-dimensional coordinate normalization.

[0037] For velocity data and acceleration data , the formula is used: ; ; is the velocity data after normalization, is the acceleration data after normalization.

[0038] For environmental data , according to its specific parameter type, the corresponding normalization formula is used for processing to obtain the normalized data set .

[0039] The process of data preprocessing is as follows: First, the original data set is input into the data cleaning algorithm module. This algorithm module judges and deletes abnormal data according to the preset reasonable range to obtain the preliminarily processed data. Then, the preliminarily processed data is input into the normalization algorithm module, and according to different data types, calculations are respectively performed according to the above normalization formulas to obtain the normalized data set .

[0040] In some embodiments, as Figure 2 shown, feature extraction: Extract effective features from the normalized data set . For position data, extract the position change trend feature by calculating the difference in position coordinates at adjacent time points, such as , , ; for velocity data, extract the velocity change rate feature, and the calculation method is ; for acceleration data, extract the acceleration change trend feature, such as ; Meanwhile, in combination with environmental data, relevant features affecting the performance of the laser galvanometer are extracted, such as the real-time value of the environmental temperature, the temperature change rate (the amount of temperature change per unit time), the temperature change trend (such as whether the temperature is rising, falling, or remaining stable), the temperature fluctuation amplitude (the difference between the maximum and minimum temperatures over a period of time), etc., and features such as the real-time value of the environmental humidity, the humidity change rate, the influence coefficient of humidity on electronic component parameters (such as resistance and capacitance values) (by experimentally determining the changes in electronic component parameters under different humidities in advance and calculating the influence coefficient), the predicted value of the corrosion degree of mechanical components caused by humidity (predicted based on an empirical model of humidity and corrosion degree or a model trained with historical data), etc. All the extracted features are combined to form a feature vector set 。

[0041] The process of feature extraction is as follows: For the normalized data set , the position, velocity, acceleration, and environmental data are respectively calculated according to the above feature extraction formula, and the various calculated features are integrated to form a feature vector set 。Meanwhile, the pendulum wave technology is introduced to further process the feature vector set . The pendulum wave technology analyzes and enhances the fluctuations of the data in the feature vector by simulating the periodic motion law of a pendulum, and mines the potential periodic features of the data to obtain an enhanced feature vector set . The specific process is as follows: Let the pendulum wave function be , where is the amplitude, is the angular frequency, is the initial phase, and is the time. The data in the feature vector set is fused and calculated with the pendulum wave function. For example, for the feature , calculate , so as to obtain an enhanced feature vector set 。

[0042] In some embodiments, the core of the application of the pendulum wave technology in the laser galvanometer control lies in modulating and enhancing the feature vector by simulating the periodic law of the pendulum motion. The following is the specific implementation process of the pendulum wave technology in step S3: The mathematical principle of the pendulum wave technology: The pendulum wave function is based on the classical pendulum motion equation, and its standard form is: ; where, is the amplitude, representing the intensity of the fluctuation (which can be dynamically adjusted according to the data characteristics); is the angular frequency, controlling the period length ( , is the period); is the initial phase, which adjusts the starting position of the waveform; is the offset, which is used for the fluctuation with a non-zero center (default = 0).

[0043] Parameter self-adaptive determination method: The parameters of the pendulum wave function need to be determined self-adaptively according to the motion characteristics of the laser galvanometer: Amplitude : ; where is the standard deviation of the feature , is the scaling factor (the empirical value is taken as 0.5 - 1.5).

[0044] Period : ; where is the average peak-to-peak time interval of the feature fluctuation, is the adjustment coefficient (the empirical value is taken as 0.8 - 1.2).

[0045] Initial phase : ;

[0046] where is the covariance, is the variance, which is used to align the change trend of the data.

[0047] Specific implementation steps of feature enhancement: Time-frequency analysis: Perform short-time Fourier transform (STFT) or wavelet transform on the original feature vector to extract the main frequency components. Identify the dominant period of the laser galvanometer movement as the initial value of the pendulum wave period.

[0048] Parameter initialization: According to the time-frequency analysis results, set the initial parameters . Use different parameter groups for environmental features (such as temperature, humidity) to decouple from the motion features.

[0049] Feature modulation: ; where is the modulation intensity coefficient (range 0.1 - 0.3), is the timestamp of the feature .

[0050] Adaptive adjustment: Every interval sampling points (such as = 100), recalculate , , . Update parameters using the data within a sliding window (e.g., window size = 50) to ensure real-time performance.

[0051] Implementation example: Assume the current processed position feature , and its time-frequency analysis shows that the dominant period is 100 ms (i.e., = 0.1). The following calculations are obtained: ; ; ; Then the pendulum wave function is: ; The enhanced feature is: .

[0052] Technical advantages: Enhanced periodic features: Highlight the periodic pattern of the galvanometer movement, improving the ability of the LSTM model to capture temporal patterns; Noise resistance: The pendulum wave function acts as a low-pass filter to suppress high-frequency noise and retain the main motion features; Self-adaptability: The parameter dynamic adjustment mechanism enables the algorithm to adapt to the galvanometer movement characteristics under different working scenarios.

[0053] For environmental features (such as temperature), a fixed-parameter pendulum wave function can be used to reduce the computational load.

[0054] In this way, the pendulum wave technology can effectively enhance the periodic features in the galvanometer movement data, improving the prediction accuracy and control stability of the subsequent model.

[0055] It should be noted that the formula focuses on directly multiplying the pendulum wave function with the original features. By simulating the periodic movement of the pendulum, potential periodic features in the data are mined; while is improved on this basis. The added term can more flexibly control the modulation intensity, avoiding feature distortion due to excessive modulation. At the same time, through the coefficient , adaptive adjustment is achieved, and data features can be better enhanced under different working conditions. From the data processing flow, the original formula serves as a basic operation to provide initially enhanced features for subsequent processing; the improved formula optimizes for possible problems in actual applications, making the feature enhancement effect more stable and reliable. For example, when the working conditions of the galvanometer are complex and variable, the improved formula can adjust the value according to different working conditions to better enhance the features and provide higher-quality data for subsequent model training.

[0056] In some embodiments, as Figure 3 shown, model training: Construct an ensemble model by combining a Long Short-Term Memory (LSTM) model with a Random Forest model. A random forest is an ensemble learning model composed of multiple decision trees, which improves the stability and prediction accuracy of the model by voting or averaging the prediction results of multiple decision trees.

[0057] Divide the enhanced feature vector set into a training set and a test set , with a division ratio of 7:3. Use the training set to train the LSTM model and the Random Forest model respectively.

[0058] For the LSTM model, during the training process, the Mean Squared Error (MSE) is used as the loss function, and its calculation formula is: ; where is the number of training samples, is the true output value of the th sample, and is the predicted output value of the th sample in the LSTM model. Continuously adjust the parameters of the LSTM model, such as the weight matrix and the bias vector , to minimize the loss function.

[0059] For the Random Forest model, during the training process, when constructing each decision tree, a part of the samples are randomly drawn from the training set with replacement (bootstrap sampling) for training. At the same time, when splitting each node, a part of the features are randomly selected to choose the optimal splitting feature. The construction process of the decision tree is based on the Gini impurity, and the Gini impurity calculation formula is: ; where represents the node, represents the number of classes, represents the node belonging to the class The sample ratio. By continuously splitting the nodes, the Gini impurity is gradually reduced until the stopping condition is met (such as the number of samples in the node is less than the threshold, the decrease in Gini impurity is less than the threshold, etc.), and the construction of a single decision tree is completed. Repeat the above process to construct multiple decision trees to form a random forest model.

[0060] When the loss value of the LSTM model on the training set converges to a small value, and the prediction error on the test set meets the preset accuracy requirements, and at the same time the prediction error of the random forest model on the test set also meets the requirements, stop the training. Integrate the LSTM model and the random forest model to obtain the final laser galvanometer motion prediction model. . The integration method is: perform weighted averaging on the prediction results of the LSTM model and the random forest model. Let the prediction result of the LSTM model be , and the prediction result of the random forest model be , and the integrated prediction result be , then , where 1 is the weighting coefficient, and the optimal value is determined by tuning parameters on the validation set.

[0061] The process of model training is as follows: First, divide the enhanced feature vector set into a training set and a test set in a ratio of 7:3. Then, input the training set into the LSTM model and the random forest model respectively. The LSTM model performs forward propagation calculation to predict the output according to the input data, and calculates the loss value according to the mean square error formula, and updates the parameters through the backpropagation algorithm and the gradient descent method; the random forest model is trained according to the method of constructing decision trees mentioned above. When both models meet the corresponding training requirements, their prediction results are weighted and integrated on the validation set, and by adjusting the weighting coefficient 1, the final model with the smallest prediction error on the validation set is obtained .

[0062] In some embodiments, real-time prediction: Input the real-time data after feature extraction into the trained laser galvanometer motion prediction model . According to the rules learned from the historical data, the model predicts the future motion state of the laser galvanometer and outputs the predicted position coordinates , speed and acceleration .

[0063] The process of real-time prediction is as follows: Input the latest feature vector extracted in step S3 into the trained model In it, the model Through forward propagation calculation, outputs the predicted future motion state data of the galvanometer .

[0064] In some embodiments, control instruction generation: According to the predicted motion state of the galvanometer And the preset target motion state , calculates the deviation between the two. The position deviation calculation formula is: ; The speed deviation calculation formula is: ; The acceleration deviation calculation formula is: ; Adopts the Proportional-Integral-Derivative (PID) control algorithm, and calculates the control instruction according to the above deviation. The output formula of the PID control algorithm is: ; Among them, Is the control instruction at time t4, Is the proportionality coefficient, Is the integral coefficient, Is the differential coefficient, Is the deviation value at time t4, Is the integral of the deviation From the initial time 0 to the current time t4 with respect to time, Is the deviation The derivative with respect to time, that is, the rate of change of the deviation at time t4. In this step, Correspond to respectively, and the control instructions for position, speed and acceleration are obtained by separate calculations , which form a control instruction set .

[0065] The process of generating control instructions is: first calculate various deviations between the predicted motion state and the target motion state, then substitute the deviation values into the PID control algorithm formula to calculate the control instructions for different motion parameters respectively, and finally integrate these control instructions into a control instruction set .

[0066] In some embodiments, control instruction execution: The control instruction set Transmitted to the control system of the galvanometer scanner. The control system adjusts the motor drive parameters of the galvanometer scanner, such as the voltage and current of the motor, according to the received control instructions, so as to realize the real-time control of the motion state of the galvanometer scanner and make the galvanometer scanner operate according to the preset target motion state.

[0067] The process of executing the control instructions is as follows: The control system receives a set of control instructions and then , , , , are respectively converted into corresponding motor drive parameter adjustment signals, and the voltage, current and other parameters of the motor are adjusted through the drive circuit, so as to control the motion of the galvanometer scanner.

[0068] In some embodiments, feedback and optimization: After the galvanometer scanner executes the control instructions, the actual motion state data is collected again through sensors, and steps S1 - S7 are repeated. The newly collected data is compared and analyzed with the predicted data and the control target data. If the deviation between the actual motion state and the target motion state exceeds the preset error range, the proportional coefficient , integral coefficient and differential coefficient in the PID control algorithm, as well as the parameters of the deep learning model, are re-adjusted to optimize the model, so as to improve the control accuracy and prediction accuracy.

[0069] The process of feedback and optimization is as follows: Collect the actual motion state data of the galvanometer scanner after executing the control instructions, perform data processing and calculation according to the process of steps S1 - S7, and compare the deviation between the actual data and the target data. If the deviation exceeds the range, according to the deviation situation, optimization algorithms (such as particle swarm optimization algorithm, genetic algorithm, etc.) are used to adjust the PID control parameters and the parameters of the deep learning model, and re-training and control are carried out until the accuracy requirements are met. At the same time, regularly evaluate the performance of the decision trees in the random forest model. If it is found that the performance of some decision trees declines, optimize them through operations such as re-training or pruning.

[0070] A galvanometer scanner intelligent control system based on big data uses the galvanometer scanner intelligent control method based on big data described in any one of the above, as Figure 4 shown. The galvanometer scanner intelligent control system based on big data includes: A data acquisition module, which is used to collect the real-time position data, motion speed data, acceleration data and environmental data of the galvanometer scanner, and respectively perform data packaging and timestamp marking processing to obtain position raw data, speed raw data, acceleration raw data and environmental raw data; A data preprocessing module for deleting abnormal data and performing normalization processing on raw position data, raw velocity data, raw acceleration data, and raw environment data to obtain valid position data, valid velocity data, valid acceleration data, and valid environment data, and integrating them to obtain a preprocessed dataset; A feature extraction module for extracting features of the changing trend of the galvanometer position, the rate of change of velocity, the changing trend of acceleration, and the relevant features of the influence of environmental factors on the performance of the galvanometer from the preprocessed dataset to form a feature vector set; using the pendulum wave technique to process the feature vector set to obtain an enhanced feature vector set; A model training module for dividing the enhanced feature vector set into a training set and a test set; training a long short-term memory network model and a random forest model respectively using the training set; performing weighted fusion on the trained long short-term memory network model and random forest model to obtain a galvanometer motion prediction model; A real-time prediction module for inputting the real-time enhanced feature vector set into the galvanometer motion prediction model to obtain the predicted position coordinates, velocity, and acceleration of the galvanometer; A control instruction generation module for calculating the position deviation, velocity deviation, and acceleration deviation between the predicted position coordinates, velocity, and acceleration of the galvanometer and the preset target motion state; adopting a proportional-integral-derivative control algorithm to calculate control instructions for position, velocity, and acceleration according to the above deviations to form a control instruction set; A control instruction execution module for transmitting the control instruction set to the control system of the galvanometer to adjust the motor drive parameters of the galvanometer and achieve real-time control of the motion state of the galvanometer; A feedback and optimization module for, after the galvanometer executes the control instructions, collecting the actual motion state data again, comparing and analyzing the newly collected data with the predicted data and the control target data. If the deviation between the actual motion state and the target motion state exceeds the preset error range, readjust the parameters in the proportional-integral-derivative control algorithm and the parameters of the galvanometer motion prediction model to optimize the model.

[0071] The above embodiments are used to illustrate the present invention, not to limit the present invention. Therefore, changes in the example values or replacement of equivalent elements should still fall within the scope of the present invention.

[0072] From the above detailed description, those of ordinary skill in the art can clearly understand that the present invention can indeed achieve the aforementioned objectives and has actually met the requirements of the patent law.

[0073] Although the preferred embodiments of the present invention have been described, those skilled in the art can make additional changes and modifications once they learn the basic creative concept. Therefore, the appended claims are intended to be construed to include the preferred embodiments as well as all changes and modifications falling within the scope of the present invention. The above description is only the preferred embodiments of the present invention and is not intended to limit the present invention. It should be noted that any modifications, equivalent replacements, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0074] It should be noted that the above description of the process is only for illustration and explanation and does not limit the scope of application of this specification. For those skilled in the art, various corrections and changes can be made to the process under the guidance of this specification. However, these corrections and changes are still within the scope of this specification.

[0075] The basic concept has been described above. Obviously, for those of ordinary skill in the art after reading this application, the above invention disclosure is only an example and does not constitute a limitation to this application. Although not explicitly stated here, those of ordinary skill in the art may make various modifications, improvements, and corrections to this application. Such modifications, improvements, and corrections are proposed in this application, so such modifications, improvements, and corrections still fall within the spirit and scope of the exemplary embodiments of this application.

[0076] At the same time, this application uses specific terms to describe the embodiments of this application. For example, "one embodiment", "an embodiment", and / or "some embodiments" mean a certain feature, structure, or characteristic related to at least one embodiment of this application. Therefore, it should be emphasized and noted that the "one embodiment" or "an embodiment" or "an alternative embodiment" mentioned two or more times in different positions in this specification does not necessarily refer to the same embodiment. In addition, certain features, structures, or characteristics in one or more embodiments of this application can be appropriately combined.

[0077] In addition, those of ordinary skill in the art can understand that various aspects of this application can be illustrated and described by several patentable types or situations, including any new and useful process, machine, product, or combination of substances, or any new and useful improvement thereof. Therefore, various aspects of this application can be implemented entirely by hardware, entirely by software (including firmware, resident software, microcode, etc.), or by a combination of hardware and software. The above hardware or software can all be referred to as "units", "modules", or "systems". In addition, various aspects of this application can take the form of a computer program product embodied in one or more computer-readable media, in which computer-readable program code is included.

[0078] The computer program code required for the operations of various parts of this application can be written in any one or more programming languages, including object-oriented programming languages such as Java, Scala, Smalltalk, Eiffel, JADE, Emerald, C++, C#, VB.NET, Python, etc., conventional procedural programming languages such as C programming language, VisualBasic, Fortran2103, Perl, COBOL2102, PHP, ABAP, dynamic programming languages such as Python, Ruby and Groovy, or other programming languages, etc. This program code can run entirely on the user's computer, or run as an independent software package on the user's computer, or partially run on the user's computer and partially run on a remote computer, or run entirely on a remote computer or server. In the latter case, the remote computer can be connected to the user's computer through any network form, such as a local area network (LAN) or a wide area network (WAN), or connected to an external computer (for example, through the Internet), or in a cloud computing environment, or used as a service such as software as a service (SaaS).

[0079] In addition, unless clearly stated in the claims, the order of the processing elements and sequences, the use of numbers and letters, or the use of other names in this application are not used to limit the order of the processes and methods of this application. Although some currently considered useful embodiments of the invention are discussed through various examples in the above disclosure, it should be understood that such details only serve the purpose of illustration, and the appended claims are not limited to the disclosed embodiments. On the contrary, the claims are intended to cover all modifications and equivalent combinations that conform to the essence and scope of the embodiments of this application. For example, although the implementation of the above various components can be embodied in a hardware device, it can also be implemented as a pure software solution, for example, installed on an existing server or mobile device.

[0080] Similarly, it should be noted that, in order to simplify the description of this application disclosure and thus help the understanding of one or more embodiments of the invention, in the previous description of the embodiments of this application, sometimes multiple features are merged into one embodiment, drawing or description thereof. However, this method of this application should not be construed as reflecting the intention that the claimed subject matter requires more features than those clearly recited in each claim. On the contrary, the subject matter of the invention should have fewer features than the above single embodiment.

Claims

1. A laser galvanometer intelligent control method based on big data, characterized in that, Including: S1. Data acquisition: Acquire the real-time position data, motion speed data, acceleration data, and environmental data of the galvanometer scanner, and perform data packing and timestamp marking processing on them respectively to obtain the original position data, original speed data, original acceleration data, and original environmental data. S2. Data preprocessing: Delete abnormal data and perform normalization processing on the original position data, original speed data, original acceleration data, and original environmental data to obtain valid position data, valid speed data, valid acceleration data, and valid environmental data, and integrate them to obtain the preprocessed dataset. S3. Feature extraction: Extract the characteristics of the position change trend of the galvanometer scanner, the speed change rate characteristics, the acceleration change trend characteristics, and the relevant characteristics of the influence of environmental factors on the performance of the galvanometer scanner from the preprocessed dataset. Form a feature vector set. Use the single pendulum wave technology to process the feature vector set to obtain an enhanced feature vector set. S4. Model training: Divide the enhanced feature vector set into a training set and a test set; use the training set to train the long short-term memory network model and the random forest model respectively; perform weighted fusion on the trained long short-term memory network model and the random forest model to obtain the galvanometer scanner motion prediction model. S5. Real-time prediction: Input the real-time enhanced feature vector set into the galvanometer scanner motion prediction model to obtain the predicted position coordinates, speed, and acceleration of the galvanometer scanner. S6. Control instruction generation: Calculate the position deviation, speed deviation, and acceleration deviation between the predicted position coordinates, speed, and acceleration of the galvanometer scanner and the preset target motion state; adopt the proportional-integral-derivative control algorithm to calculate the control instructions for position, speed, and acceleration according to the above deviations, and form a control instruction set. S7. Control instruction execution: Transmit the control instruction set to the control system of the galvanometer scanner, adjust the motor drive parameters of the galvanometer scanner, and realize the real-time control of the motion state of the galvanometer scanner. S8. Feedback and optimization: After the galvanometer scanner executes the control instruction, collect its actual motion state data again, and repeat the above steps S1 - S7; compare and analyze the newly collected data with the predicted data and the control target data. If the deviation between the actual motion state and the target motion state exceeds the preset error range, readjust the parameters in the proportional-integral-derivative control algorithm and the parameters of the galvanometer scanner motion prediction model to optimize the model.

2. The intelligent control method of a laser galvanometer based on big data according to claim 1, characterized in that In S1, set a fixed sampling frequency to periodically collect the position data, speed data, acceleration data, and environmental data of the galvanometer scanner.

3. The intelligent control method for a laser galvanometer based on big data according to claim 1, wherein In S2, the judgment basis for abnormal data is that the position coordinates, speed, acceleration, or environmental data of the galvanometer scanner exceed the preset reasonable range.

4. The intelligent control method of a laser galvanometer based on big data according to claim 1, wherein In S3, the position change trend characteristic is obtained by calculating the difference between the position coordinates of adjacent time points; the speed change rate characteristic is obtained by calculating the ratio of the difference between the speeds of adjacent time points to the time interval; the acceleration change trend characteristic is obtained by calculating the difference between the accelerations of adjacent time points.

5. The intelligent control method of a laser galvanometer based on big data according to claim 1, characterized in that In S3, the pendulum wave technique performs fluctuation analysis and enhancement on the data in the eigenvector set by simulating the periodic motion law of a pendulum, and mines the potential periodic features of the data.

6. The intelligent control method of a laser galvanometer based on big data according to claim 1, wherein In S4, during the training process of the long short-term memory network model, the mean square error is used as the loss function, and the model parameters are adjusted through the backpropagation algorithm.

7. The intelligent control method of a laser galvanometer based on big data according to claim 1, characterized in that In S4, the random forest model consists of multiple decision trees. When each decision tree is constructed, training samples are randomly selected with replacement, and some features are randomly selected for node splitting.

8. The intelligent control method of a laser galvanometer based on big data according to claim 1, characterized in that In S4, when the long short-term memory network model and the random forest model are weighted and fused, the weighting coefficients are determined by tuning parameters on the validation set.

9. The intelligent control method of a laser galvanometer based on big data according to claim 1, characterized in that In S8, the particle swarm optimization algorithm or the genetic algorithm is used to adjust the parameters in the proportional-integral-derivative control algorithm and the parameters of the laser galvanometer motion prediction model.

10. An intelligent control system for a laser galvanometer based on big data, characterized in that, Using the big data-based intelligent control method for a laser galvanometer according to any one of claims 1 to 9, the big data-based intelligent control system for a laser galvanometer includes: A data acquisition module, configured to acquire the real-time position data, motion speed data, acceleration data, and environmental data of the laser galvanometer, perform data packing and timestamp marking processing respectively, and obtain the original position data, original speed data, original acceleration data, and original environmental data; A data preprocessing module, configured to delete abnormal data and perform normalization processing on the original position data, original speed data, original acceleration data, and original environmental data, obtain the valid position data, valid speed data, valid acceleration data, and valid environmental data, and integrate them to obtain a preprocessed data set; A feature extraction module, configured to extract the laser galvanometer position change trend features, speed change rate features, acceleration change trend features, and relevant features of the influence of environmental factors on the laser galvanometer performance from the preprocessed data set to form an eigenvector set; use the pendulum wave technique to process the eigenvector set to obtain an enhanced eigenvector set; A model training module, configured to divide the enhanced eigenvector set into a training set and a test set; use the training set to train the long short-term memory network model and the random forest model respectively; perform weighted fusion on the trained long short-term memory network model and the random forest model to obtain a laser galvanometer motion prediction model; A real-time prediction module, configured to input the real-time enhanced eigenvector set into the laser galvanometer motion prediction model to obtain the predicted laser galvanometer position coordinates, speed, and acceleration; A control instruction generation module, configured to calculate the position deviation, speed deviation, and acceleration deviation between the predicted laser galvanometer position coordinates, speed, and acceleration and the preset target motion state; adopt the proportional-integral-derivative control algorithm to calculate the control instructions for the position, speed, and acceleration according to the above deviations to form a control instruction set; A control instruction execution module, configured to transmit the control instruction set to the control system of the laser galvanometer, adjust the motor drive parameters of the laser galvanometer, and realize the real-time control of the motion state of the laser galvanometer; A feedback and optimization module is used to collect the actual motion state data of the galvanometer scanner again after it executes the control instruction. The newly collected data is compared and analyzed with the predicted data and the control target data. If the deviation between the actual motion state and the target motion state exceeds the preset error range, the parameters in the proportional-integral-derivative control algorithm and the parameters of the motion prediction model of the galvanometer scanner are readjusted to optimize the model.

Images