Intelligent control method, system and device for vacuum isothermal forging press

Abstract

The invention relates to the technical field of forging process intelligent control, in particular to an intelligent control method, system and device for a vacuum isothermal forging press. The method comprises the steps that temperature, pressure and stress data in the forging process are collected based on a multi-mode sensor, and a preliminary original data sequence is generated; cleaning and standardizing the original data sequence to generate standardized real-time working condition data; based on the real-time working condition data and the historical data, constructing a dynamic process prediction model, and generating a key parameter prediction sequence; inputting the key parameter prediction sequence into a reinforcement learning model to generate an optimized process parameter set; and dynamically adjusting a press execution system based on the optimized process parameter set, monitoring the equipment operation state in real time, and generating a closed-loop control instruction sequence. The forging quality consistency and the production efficiency can be effectively improved, and the equipment failure risk is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of forging process intelligent control, and in particular to a vacuum isothermal forging press intelligent control method, system and device. Background Art

[0002] Vacuum isothermal forging is a metal forming process performed under high temperature and low pressure conditions. This process can significantly improve the internal microstructure uniformity and material properties of forgings by isothermally deforming metal materials for a long time in a vacuum environment. However, due to the dynamic coupling characteristics of process parameters such as temperature, pressure and press speed during the forging process, traditional control methods are difficult to meet the high requirements for process accuracy and stability.

[0003] In the prior art, forging presses are usually controlled by preset fixed process parameters, which cannot respond to the dynamic changes of multiple variables in the forging process in real time. In addition, the monitoring and fault warning mechanisms of the equipment operation status are not perfect, which can easily lead to unstable quality of forged products and increased equipment downtime.

[0004] With the rapid development of artificial intelligence technology, the application of deep learning and reinforcement learning algorithms in complex industrial scenarios has gradually matured. However, applying them to the dynamic optimization control of vacuum isothermal forging process still faces many challenges, such as how to extract key process features from multimodal sensor data, how to establish accurate dynamic prediction models based on real-time and historical data, and how to use reinforcement learning to achieve intelligent optimization of process parameters.

[0005] Therefore, in view of the shortcomings of existing vacuum isothermal forging presses in process parameter optimization, equipment status monitoring and fault warning, there is an urgent need for an intelligent control method and system that can combine multimodal sensor data, deep learning and reinforcement learning algorithms to achieve dynamic optimization control of process parameters and efficient monitoring of equipment operating status. This innovative technology can not only improve the consistency and reliability of forging quality, but also significantly reduce the risk of equipment failure and improve production efficiency. Summary of the invention

[0006] The present invention provides an intelligent control method, system and device for a vacuum isothermal forging press to solve the problem of how to dynamically optimize process parameters such as temperature, pressure and press speed during the forging process based on real-time operating data and historical data collected by multimodal sensors, combined with deep learning and reinforcement learning algorithms, monitor the equipment operating status in real time, generate closed-loop control instructions, improve the quality of forgings and reduce the risk of equipment failure.

[0007] In order to solve the above technical problems, the present invention provides an intelligent control method for a vacuum isothermal forging press, comprising: Based on multimodal sensors, the temperature, pressure and stress data of the forging process are collected to generate a preliminary raw data sequence; Cleaning and standardizing the original data sequence to generate standardized real-time operating condition data; Based on the real-time operating data and historical data, a deep learning model is used to extract key process features, a dynamic process prediction model is constructed, and a key parameter prediction sequence is generated; Inputting the key parameter prediction sequence into a reinforcement learning model, optimizing process parameters such as temperature, pressure and press speed during the forging process, and generating an optimized process parameter set; Based on the optimized process parameter set, the press execution system is dynamically adjusted, and the equipment operation status is monitored in real time to generate a closed-loop control instruction sequence.

[0008] Furthermore, the collecting of temperature, pressure and stress data during the forging process includes: Acquire temperature data through thermocouples and infrared detectors; Obtain pressure data through strain gauges; The stress data is obtained through the mechanical sensor.

[0009] Furthermore, the cleaning and standardization of the original data sequence specifically includes: Using Kalman filtering algorithm to filter noise from the original data sequence; Detect and remove outliers to obtain cleaned data; The cleaned data is standardized based on the z-score standardization method to generate standardized real-time operating condition data.

[0010] Furthermore, the extraction of key process features specifically includes: Based on the standardized real-time operating data, Fourier transform is used to extract key features of temperature change rate, pressure fluctuation amplitude and stress response; The extracted features are screened using the principal component analysis method to generate a process feature data set.

[0011] Furthermore, the construction of a dynamic process prediction model specifically includes: Based on the process feature data set and historical process data, a dynamic process prediction model is established using a recurrent neural network model; The process feature data set is input, key parameters in the forging process are predicted, and a key parameter prediction sequence is generated.

[0012] Furthermore, the optimized forging process parameters specifically include: Using the key parameter prediction sequence as the state input of the reinforcement learning model; Optimize the process parameter adjustment strategy through reinforcement learning algorithm to generate the optimal process parameter combination; Output optimized temperature, pressure and press speed parameters.

[0013] Furthermore, the closed-loop control instruction generation specifically includes: Collect real-time operating status data of the equipment, including real-time temperature, pressure and press speed; generating a closed-loop control instruction sequence based on the feedback data and the optimized process parameter set; The temperature, pressure and press speed are continuously adjusted during the forging process.

[0014] Furthermore, the vacuum isothermal forging press intelligent control method further includes: Use time series prediction models to analyze equipment operating status and generate fault warning information; Based on the fault warning information, the process parameters are dynamically corrected.

[0015] Furthermore, a vacuum isothermal forging press intelligent control system is provided, the system comprising: A data acquisition module is used to collect temperature, pressure and stress data during the forging process from multimodal sensors to generate a raw data sequence; A data processing module, used for cleaning and standardizing the raw data sequence to generate standardized real-time operating condition data; A dynamic process prediction module is used to extract key process characteristics, build a dynamic process prediction model, and generate a key parameter prediction sequence based on the real-time operating data and historical data; A process parameter optimization module, used to optimize the process parameters based on the key parameter prediction sequence using a reinforcement learning model to generate an optimized process parameter set; The closed-loop control module is used to generate a closed-loop control instruction sequence based on the optimized process parameter set to dynamically adjust the press execution system.

[0016] Furthermore, a vacuum isothermal forging press intelligent control device includes a data acquisition module, a data processing and feature extraction module, a dynamic process prediction module, a process parameter optimization module, and a closed-loop control module. The device can implement the steps of the above-mentioned vacuum isothermal forging press intelligent control method.

[0017] The key innovative features of the present invention include: (1) Data acquisition and processing technology based on multimodal sensors. Innovatively combining multimodal sensors with data cleaning and standardization to generate high-precision real-time working condition data, effectively avoiding process parameter deviations caused by data noise or outliers.

[0018] (2) Construction of dynamic process prediction model. Use deep learning algorithms (such as recurrent neural networks, RNNs) to model real-time operating data, combine historical data and prior knowledge, accurately predict the changing trends of key parameters, and solve the nonlinear coupling problem under complex operating conditions.

[0019] (3) Process parameter optimization based on reinforcement learning. The predicted parameters are dynamically adjusted through reinforcement learning algorithms to optimize the process parameter combination, thereby improving the adaptability and optimization efficiency of equipment and processes in the production process.

[0020] The following are its main beneficial effects: (1) High-precision real-time operating data collection and processing. The present invention collects multi-dimensional real-time operating data such as temperature, pressure and stress based on multimodal sensors, and generates standardized high-quality operating data through noise filtering and outlier removal technology. Compared with the traditional fixed parameter setting method, the real-time collected and processed data is more accurate and comprehensive, laying the foundation for subsequent dynamic optimization.

[0021] (2) Dynamic prediction and optimization improve process accuracy. The deep learning model is used to extract key process features, and a dynamic process prediction model is constructed in combination with historical data and expert knowledge, which can accurately predict the changing trend of future process parameters. The reinforcement learning model is used to optimize parameters such as temperature, pressure, and press speed during the forging process, solving the problem that traditional methods are difficult to adapt to complex dynamic working conditions, and greatly improving the consistency of forging quality.

[0022] (3) Closed-loop control reduces the risk of equipment failure. Based on real-time monitoring of the equipment's operating status and generating a closed-loop control instruction sequence, the equipment's operating parameters can be dynamically adjusted to respond to equipment abnormalities in real time, reducing the probability of equipment failure. Compared with traditional manual intervention or fixed rule control methods, this invention significantly improves the continuity and reliability of production. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 A schematic diagram of a flow chart of an intelligent control method for a vacuum isothermal forging press provided in an embodiment of the present application; Figure 2 A structural block diagram of the intelligent control system of the vacuum isothermal forging press provided in the embodiment of the present application; Figure 3 This is a structural block diagram of the intelligent control device for a vacuum isothermal forging press provided in an embodiment of the present application. DETAILED DESCRIPTION

[0024] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by technicians in the technical field of the present application; the terms used in the specification of the application herein are only for the purpose of describing specific embodiments and are not intended to limit the present application; the terms "including" and "having" and any variations thereof in the specification and claims of the present application and the above-mentioned drawings are intended to cover non-exclusive inclusions. The terms "first", "second", etc. in the specification and claims of the present application or the above-mentioned drawings are used to distinguish different objects, not to describe a specific order.

[0025] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0026] Example 1: Reference Figure 1 , is a flow chart of a vacuum isothermal forging press intelligent control method provided by an embodiment of the present invention, and the flow chart may at least include steps S100-S500: S100, collecting temperature, pressure and stress data during the forging process based on a multimodal sensor to generate a preliminary raw data sequence; S200, cleaning and standardizing the original data sequence to generate standardized real-time operating condition data; S300, based on the real-time operating data and historical data, using a deep learning model to extract key process features, build a dynamic process prediction model, and generate a key parameter prediction sequence; S400, inputting the key parameter prediction sequence into a reinforcement learning model, optimizing process parameters such as temperature, pressure and press speed during forging, and generating an optimized process parameter set; S500: Based on the optimized process parameter set, dynamically adjust the press execution system, monitor the equipment operation status in real time, and generate a closed-loop control instruction sequence.

[0027] Step S100 at least includes steps S110-S130: S110, obtain the temperature, pressure and stress data during the forging process, use the multimodal sensor to monitor the process parameters in real time, and generate a preliminary raw data sequence through the sensor data acquisition module. Sensor data acquisition: During the operation of the vacuum isothermal forging press, the multimodal sensor is arranged at the key position of the forging system to monitor the process parameters such as temperature, pressure and stress in real time. Specifically, the temperature sensor measures the temperature of the press cavity through a thermocouple or an infrared detector, the pressure sensor monitors the pressure change in the press through a strain gauge or a piezoelectric sensor, and the stress sensor records the stress generated during the forging process through a mechanical sensor.

[0028] Data acquisition module: The raw signals collected by the sensor are transmitted in real time through the data acquisition module. Specifically, the data acquisition module includes a signal processing unit, which performs pre-processing such as conversion, amplification, and filtering on the analog signals from different sensors so as to convert them into digital signals. The data of each sensor is processed synchronously through time series, and multi-dimensional data such as temperature, pressure, and stress are aggregated into a preliminary raw data sequence (hereinafter referred to as the raw data sequence).

[0029] Assuming temperature is T, pressure is P, and stress is σ, the initial raw data sequence is , where the value of each parameter changes over time, Includes temperature, pressure and stress data at every moment in the forging process.

[0030] Real-time data recording: The raw data sequence will be recorded in real time and provide basic input data for subsequent data cleaning and processing.

[0031] S120: Noise filtering and outlier elimination are performed on the original data sequence, and data standardization is performed to generate standardized working condition data. Noise filtering and outlier removal: The original data sequence Contains certain noise and abnormal values ​​(such as abnormal data caused by sensor failure or environmental interference). First, Kalman filtering or wavelet denoising algorithm is used to De-noising is performed to eliminate random noise introduced by sampling errors in the data. Specifically, Kalman filtering minimizes the impact of noise by predicting and updating the signal, thereby obtaining more accurate temperature, pressure and stress data.

[0032] Outlier removal: Further, outlier detection is performed on the denoised data. By calculating the statistical characteristics (such as mean, variance, etc.) of each parameter (temperature T, pressure P, and stress σ), and setting a reasonable threshold (for example, the outlier standard is mean ± 3 times standard deviation), the values ​​that are out of the normal range are screened out and removed from the data sequence.

[0033] The normal range of set temperature T, pressure P, and stress σ is: By removing abnormal data outside these ranges, we can obtain standardized data without noise and outliers. .

[0034] Data standardization: In order to standardize the data and eliminate the errors caused by different dimensions, the z-score standardization method is used. Specifically, the mean μ and standard deviation σ of each parameter are calculated and standardized: in, is the standardized data, μ is the mean of temperature, pressure and stress, and σ is the standard deviation. This step will generate a standardized operating data sequence , each of whose parameters has a mean of 0 and a standard deviation of 1.

[0035] Output standardized data: the standardized working condition data It will be passed as input data to the subsequent feature extraction stage (i.e., S130 module).

[0036] S130, extract key features from the standardized working condition data sequence, perform feature selection and feature engineering, and obtain high-quality real-time working condition data Feature extraction: normalized working condition data On this basis, signal processing and data analysis methods are used to extract key features. Specifically, Fourier transform, waveform analysis and other technologies are used to perform frequency domain and time domain analysis on the time series data of temperature T, pressure P and stress σ to extract key characteristic parameters such as frequency component, amplitude, phase, etc. In addition, the statistical features such as the rate of change, maximum value, minimum value, mean value, etc. of each parameter can be extracted in combination with dynamic characteristics.

[0037] Feature selection: From the extracted multidimensional features, select the features that are more relevant to the forging process quality. Use methods such as principal component analysis (PCA), mutual information method, correlation analysis, etc. to screen out important features that can effectively reflect the forging process. For example, in some cases, the temperature change rate is highly correlated with the material deformation quality, while changes in pressure and stress may have a greater impact on the dimensional accuracy of the finished product.

[0038] Suppose the extracted features are , these features are used for subsequent process modeling.

[0039] Feature engineering: Based on the selected features, further feature engineering is performed. In combination with the needs of machine learning and deep learning, the original features are fused, transformed, or derived into new features. For example, the combined features of temperature, pressure, and stress may be weighted averaged or synthesized based on historical data and prior knowledge to provide more accurate input for subsequent modeling.

[0040] Generate high-quality real-time operating data: Finally, after feature extraction and selection, high-quality real-time operating data is obtained , which will be used as input data in the subsequent dynamic process prediction model (S200 module).

[0041] In the implementation of modules S110, S120, and S130, the temperature, pressure, and stress data of the forging process are obtained in real time through multimodal sensors (S110 module), and these data are subjected to noise filtering, outlier elimination, and standardization (S120 module), and finally the key features that have the greatest impact on the forging process are extracted (S130 module). The output data of these steps will serve as crucial input data in subsequent deep learning and reinforcement learning algorithms to build dynamic process prediction models and optimize forging process parameters.

[0042] Step S200 at least includes steps S210-S230: S210. Based on the real-time working condition data and historical data, a deep learning model is used to extract key process features in the forging process and construct a process feature data set.

[0043] Data input: the real-time working condition data (Standardized data of temperature T, pressure P, stress σ) and historical data (Past forging process data, including historical temperature, pressure, stress and other parameters) will be used as input data for the deep learning model. Specifically, real-time working condition data is combined with historical data to form a multi-dimensional input data set. .

[0044] Deep learning model training: Use deep learning models such as deep neural network (DNN), convolutional neural network (CNN) or long short-term memory network (LSTM) to train the data set Through training, the deep learning model can extract important process features of the forging process. Specifically, the model gradually adjusts the network parameters through the forward propagation and back propagation mechanisms, so that the model can effectively learn the key factors that affect the forging quality and process stability.

[0045] Process feature extraction: The deep learning model is trained to automatically extract process features from input data. The key process features that help predict the quality and efficiency of the forging process are extracted from the forging process. Specifically, the extracted features can include temperature change rate, pressure fluctuation amplitude, stress response, etc. These features play an important role in optimizing the forging process.

[0046] Output process feature dataset: The key process features extracted by the deep learning model constitute the process feature dataset ,This dataset will be used for subsequent process modeling and prediction.

[0047] in, The key process characteristics of temperature, pressure and stress are represented respectively. This data set provides input for the subsequent dynamic process prediction model.

[0048] S220. Combining expert knowledge and prior information, the process feature data is integrated with the historical process data, a dynamic process prediction model is constructed, process modeling is performed, and a process parameter prediction sequence is obtained.

[0049] Introduction of expert knowledge and prior information: Combine expert knowledge and industry experience in the forging field to further analyze process feature data Specifically, the learning process of the model is guided by incorporating prior knowledge, such as the process range of specific temperature, pressure and stress, or the law of process parameter changes in historical experience. For example, experts may already know the deformation behavior of a certain material under specific pressure and temperature, so this knowledge can be incorporated into the model as prior information.

[0050] Integration of historical process data: (including process parameters of past forging processes and their corresponding quality results) and process characteristic data Fusion. Through data fusion, the prediction ability of the model is further improved. The historical process data can be combined with the current real-time working condition data using methods such as weighted average, splicing or additive model to obtain a fused data set. .

[0051] Building a dynamic process prediction model: Leveraging fused datasets , using time series models such as recurrent neural network (RNN) or long short-term memory network (LSTM) in deep learning to build a dynamic process prediction model. This model can not only predict the changes in key parameters in the forging process based on the current process characteristics, but also capture the time series correlation in the process and predict the trend of process parameter changes in a certain period of time in the future.

[0052] Set the model output to , that is, the future predicted values ​​of process parameters such as temperature, pressure, and stress during the forging process.

[0053] Output process parameter prediction sequence: After model training, the dynamic process prediction model will output the predicted process parameter sequence , this sequence will describe the changing trends of parameters such as temperature, pressure, and stress during the forging process in the future time window.

[0054] in, The predicted temperature, pressure and stress parameters represent the changes in process parameters in the future time period.

[0055] S230, based on the dynamic process prediction model and the extracted process features, generate future prediction values ​​of parameters such as temperature, pressure, stress, etc. in the forging process, and provide an adjustable parameter prediction range.

[0056] Generate future values ​​based on prediction model: The dynamic process prediction model is based on the input process characteristic data And the fused data Based on the above, the future prediction values ​​of temperature, pressure, stress and other parameters in the forging process are generated. Specifically, the predicted value is based on the temporal characteristics of the model and can reflect the trend of the change of each parameter over time during the forging process.

[0057] Adjustment and control of prediction range: According to the feedback of real-time working condition data and the process range set by experts, adjust the parameter range predicted by the model. For example, the temperature range can be set to 900℃ to 1200℃ and the pressure range can be set to 50MPa to 100MPa based on expert knowledge. According to these settings, the predicted value output by the dynamic process prediction model will be automatically adjusted within the set range, thereby ensuring the rationality of the prediction and matching the actual production conditions.

[0058] Set the adjusted temperature range to , pressure range is , the stress range is .

[0059] Output predicted values ​​and adjustment ranges: Ultimately, the dynamic process prediction model will generate future predicted values ​​for the forging process , and combined with the adjustment range set by experts, provide an adjustable parameter prediction range. The output value will be used for subsequent process optimization and closed-loop control to ensure that the forging process is carried out within a reasonable parameter range.

[0060] in, The predicted values ​​are adjusted to ensure that each process parameter meets the control requirements of the forging process.

[0061] In the S210, S220 and S230 modules, based on real-time operating data , extract key process features through deep learning models, build dynamic process prediction models, and optimize them in combination with expert knowledge, ultimately generating future prediction values ​​for parameters such as temperature, pressure, and stress during the forging process. These prediction values ​​and their adjustable prediction ranges will be used for subsequent process optimization and real-time adjustments to ensure the stability and quality of the forging process.

[0062] Step S300 at least includes steps S310-S330: S310: Inputting the key process parameter prediction sequence into the reinforcement learning model, and dynamically adjusting it through the reinforcement learning algorithm to obtain the optimal process parameter combination.

[0063] Reinforcement learning model construction: The reinforcement learning model consists of a state space, an action space, and a reward function. The state space is composed of the input process parameter sequence, including the predicted values ​​from the S220 and S230 modules (i.e. ). In this step, the reinforcement learning model decides how to adjust the process parameters to optimize the forging process by observing the current state S(t) (i.e., the current temperature, pressure, stress, etc.).

[0064] Definition of state space and action space: Specifically, the state space S(t) can be expressed as: Among them, T(t), P(t), σ(t) are the actual values ​​of temperature, pressure and stress at the current time t. The action space A(t) is a set of adjustable process parameters, usually including temperature .

[0065] Dynamic adjustment and strategy optimization: The reinforcement learning algorithm uses the strategy π(S(t)) to calculate the optimal action a(t) based on the current state S(t), that is, to adjust the specific values ​​of temperature, pressure and press speed. The specific control strategy is: The reinforcement learning algorithm optimizes π(S(t)) through a continuous training process to find an optimal parameter combination that can be continuously improved over multiple attempts. The optimization goal is to maximize a certain reward function (for example, improving process stability, energy saving, or product quality).

[0066] Optimal process parameter combination output: The reinforcement learning algorithm outputs the optimal process parameter combination according to the strategy , that is, the specific values ​​of the adjusted temperature, pressure and press speed, which represent the most suitable process parameters in the forging process.

[0067] in, They respectively represent the temperature, pressure and press speed after reinforcement learning optimization at time t.

[0068] S320: Based on the optimization solution output by the reinforcement learning model, adjust process parameters such as temperature, pressure, and press speed to obtain an optimal set of process parameters.

[0069] Optimization solution application: based on the optimal process parameter combination output by the S310 module , adjust the control system parameters in the actual forging process. Specifically, the control system will optimize the temperature ,pressure and press speed Input to the forging press system.

[0070] Temperature, pressure and press speed regulation: The control system dynamically adjusts the equipment’s operating parameters based on the output optimized by the reinforcement learning algorithm: Real-time monitoring and feedback: During the adjustment process, the control system monitors process parameters such as temperature, pressure, and press speed in real time to ensure that the actual value is consistent with the optimized value. If the deviation exceeds the predetermined range, it will be readjusted. The real-time feedback system feeds this information back to the reinforcement learning algorithm so that it can further optimize the control strategy.

[0071] S330: Performing accuracy verification and parameter debugging on the optimized process parameter set to generate a final process parameter set.

[0072] Precision verification: After the S320 module optimizes the process parameters, the system will verify the accuracy of key process parameters such as temperature, pressure, press speed, etc. By comparing with the actual production results, it is evaluated whether the optimized process parameters can achieve the expected results.

[0073] For example, verify the quality of the product during the forging process and check whether there are problems such as uneven temperatures, unstable pressure or deviations in the press speed.

[0074] At this time, the predicted and actual values ​​of temperature, pressure and press speed should be consistent. If the deviation is large, the optimization parameters need to be debugged.

[0075] Parameter debugging: During the accuracy verification process, if some process parameters are found to be deviated or do not meet the production requirements, corresponding parameter debugging is performed. For example, the temperature range may need to be adjusted. Make fine adjustments or adjust the pressure fluctuation range , so that it meets the product requirements.

[0076] Generate the final process parameter set: Through precision verification and parameter debugging, the final process parameter set after precise tuning is generated , which represents the final control parameters required in the actual forging process.

[0077] in, They are the temperature, pressure and press speed parameters after accuracy verification and debugging.

[0078] In the S310, S320 and S330 modules, the process parameters are continuously optimized through the reinforcement learning model, and the optimal process parameter set is finally generated. Among them, the reinforcement learning algorithm in S310 predicts the sequence based on key process parameters Dynamically adjust parameters. In S320, the output optimization solution is applied to the forging control system for parameter adjustment. In S330, the optimized process parameters are verified and debugged to generate the final process parameter set. This process ensures the optimization and stability of the forging process.

[0079] Step S400 at least includes steps S410-S430: S410: Based on the optimized process parameter set, a process control instruction is input to the press execution system to adjust the working state of the press in real time.

[0080] Input control instructions: The control instruction sequence C(t)C(t)C(t) generated in the S400 module will be input into the press execution system in real time. Specifically, the instruction sequence includes the target temperature at time t , Target Pressure and target press speed The execution system adjusts the state of the press according to these inputs. At this point, the control instructions can be expressed in the following ways: These parameters dynamically adjust the equipment to ensure the press system operates as planned.

[0081] Press execution system response: According to the input control instructions, the press execution system will adjust the working state in real time, including the synchronous adjustment of the heating system, pressure regulation system and press speed regulation system. The temperature, pressure and speed will be adjusted quickly under the command of the control system to achieve the optimal process conditions.

[0082] Feedback status monitoring: At the same time, the press execution system monitors the current temperature T(t), pressure P(t) and press speed v(t), and feeds the monitoring data back to the control system to ensure that the equipment is adjusted according to the target process parameters.

[0083] S420: Monitor the operating status of the press in real time, and feed the monitoring data back to the closed-loop control system to dynamically adjust the control parameters.

[0084] Operation status monitoring: After the control command input in the S410 module drives the press execution system to run, the system starts to monitor the working status of the press in real time. Specifically, the monitoring data includes the real-time temperature T(t), real-time pressure P(t) and real-time press speed v(t) of the press, which are collected by sensors in real time.

[0085] Condition monitoring feedback: The temperature, pressure and press speed obtained from real-time monitoring will be input into the closed-loop control system as feedback data F(t). These feedback data can be expressed as: in, , , are the temperature, pressure and press speed measured by the sensor at time t respectively.

[0086] Dynamic adjustment of control parameters: Based on the feedback data F(t), the control system will dynamically adjust the current temperature T(t), pressure P(t) and press speed v(t). If there is a deviation between the feedback data and the target value, the system will correct the control parameters based on the set control algorithm (such as proportional-integral-differential control).

[0087] Specifically, if the deviation of temperature, pressure or speed is too large, the control system will adjust the corresponding control instructions in real time to reduce the deviation and ensure that the equipment remains in an optimized working state. These adjustments can be expressed by the control parameter adjustment formula as follows: Among them, ΔC(t) represents the control command after dynamic adjustment according to the deviation.

[0088] S430: Generate a closed-loop control instruction sequence based on the real-time feedback data to continuously adjust key process parameters such as temperature, pressure, and speed.

[0089] Processing of real-time feedback data: Based on the real-time feedback data F(t) obtained in the S420 module, the control system will analyze the operating status of the press in real time and evaluate the deviation. According to the control strategy, the control system will calculate new control instructions. , in order to achieve continuous optimization of process parameters.

[0090] Generate closed-loop control instruction sequence: In the S430 module, based on the feedback data, the control system will regenerate the closed-loop control instruction sequence Cloop(t)C_{loop}(t)Cloop(t), which represents the new target values ​​of temperature, pressure and press speed. The specific control instruction can be expressed as: in, The temperature, pressure and press speed are adjusted according to the feedback data.

[0091] Continuously adjust process parameters: Generate closed-loop control instruction sequence After that, the system will continue to adjust key process parameters such as temperature, pressure, and press speed according to this sequence to keep the equipment running in the optimal state. This process is a dynamic adjustment process, ensuring that the equipment can always respond to environmental changes and correct operating parameters in real time.

[0092] Step S500 at least includes steps S510-S530: S510: Based on the real-time operating condition data and operating status sequence, a time series prediction model is used to analyze the operating status of the equipment to identify potential failure risks.

[0093] Fault risk identification: In the S500 module, the timing prediction model Provides predictions of the future state of the equipment By comparing the actual operation data The difference between the actual value and the predicted value can identify the potential failure risk of the equipment. Specifically, the fault identification algorithm is used to determine the potential failure risk based on the equipment's operating deviation ΔT(t), ΔP(t), and Δv(t). For example: Among them, R(t) is the failure risk value of the equipment at time t, and f(⋅) represents the fault identification function. This function can be a simple weighted sum or dynamically adjusted according to the risk level of different process parameters.

[0094] Analysis of prediction results: Further process the analysis results of R(t) and output the probability of failure. If R(t) reaches a certain threshold , it means that the device has a potential failure risk and needs further processing.

[0095] S520: Perform anomaly detection and fault identification on the prediction result, generate fault warning information, and provide reference for equipment fault repair.

[0096] Anomaly detection and fault identification: Use the fault risk value R(t) generated in S510 to detect anomalies in the equipment operation status. By detecting the deviation between the real-time operation status F(t) and the predicted status, it is determined whether the equipment is abnormal. The anomaly detection method can determine whether there is an anomaly based on the set threshold: Among them, ϵ is the set threshold. When ΔR(t) exceeds the threshold, it indicates that the equipment is faulty or abnormal.

[0097] Generate fault warning information: When the failure risk of the equipment exceeds the set threshold R_{th}, the system will generate fault warning information W(t). Specifically, the fault warning information can be expressed as: in, They are fault warning information of temperature, pressure and press speed respectively.

[0098] Fault repair reference: The generated fault warning information provides a reference for equipment maintenance personnel to help identify the possible source of equipment failure and provide support for subsequent fault repair work. According to different fault types, corresponding repair plans are formulated.

[0099] S530: Based on the fault warning information, the process parameters are adjusted using an adaptive strategy or an automatic repair mode is started to dynamically correct the process parameters.

[0100] Adaptive strategy to adjust process parameters: Based on the fault warning information W(t) generated in S520, the system can automatically adjust the process parameters or enable the repair mode. Specifically, if the fault warning signal indicates that a process parameter (such as temperature, pressure or press speed) is abnormal, the system will adjust the corresponding process parameter through the adaptive strategy. For example, if Indicates abnormal temperature, the system will automatically adjust the target temperature To correct the temperature deviation, the specific adjustment formula is as follows: in, It is the temperature correction value that is dynamically adjusted according to the fault warning information.

[0101] Start the automatic repair mode: In some cases, the fault warning information W(t) indicates that the equipment has a serious fault, and the system will automatically start the repair mode. The automatic repair mode may include stopping the equipment operation, starting emergency cooling or other repair measures. The specific repair mode can be determined by the fault type and the corresponding repair operation is started.

[0102] Dynamic correction of process parameters: Based on fault warning information and adaptive strategy adjustment, the system will dynamically correct process parameters to ensure that the equipment can resume normal operation. The dynamic correction process includes adjusting the temperature T(t), pressure P(t) and press speed v(t) to restore the equipment to the optimal working state.

[0103] In modules S510, S520, and S530, the operating status of the equipment is first analyzed through the time series prediction model to identify potential failure risks, and abnormality detection and fault identification are performed based on the prediction results. Then, based on the fault warning information, the system adjusts the process parameters or starts the automatic repair mode through an adaptive strategy to ensure that the equipment can quickly recover to the optimal working condition when a failure occurs. All steps are closely connected and rely on real-time data and feedback information obtained in previous modules (such as S420).

[0104] The key innovative features of the present invention include: (1) Data acquisition and processing technology based on multimodal sensors. Innovatively combining multimodal sensors with data cleaning and standardization to generate high-precision real-time working condition data, effectively avoiding process parameter deviations caused by data noise or outliers.

[0105] (2) Construction of dynamic process prediction model. Use deep learning algorithms (such as recurrent neural networks, RNNs) to model real-time operating data, combine historical data and prior knowledge, accurately predict the changing trends of key parameters, and solve the nonlinear coupling problem under complex operating conditions.

[0106] (3) Process parameter optimization based on reinforcement learning. The predicted parameters are dynamically adjusted through reinforcement learning algorithms to optimize the process parameter combination, thereby improving the adaptability and optimization efficiency of equipment and processes in the production process.

[0107] The following are its main beneficial effects: (1) High-precision real-time operating data collection and processing. The present invention collects multi-dimensional real-time operating data such as temperature, pressure and stress based on multimodal sensors, and generates standardized high-quality operating data through noise filtering and outlier removal technology. Compared with the traditional fixed parameter setting method, the real-time collected and processed data is more accurate and comprehensive, laying the foundation for subsequent dynamic optimization.

[0108] (2) Dynamic prediction and optimization improve process accuracy. The deep learning model is used to extract key process features, and a dynamic process prediction model is constructed in combination with historical data and expert knowledge, which can accurately predict the changing trend of future process parameters. The reinforcement learning model is used to optimize parameters such as temperature, pressure, and press speed during the forging process, solving the problem that traditional methods are difficult to adapt to complex dynamic working conditions, and greatly improving the consistency of forging quality.

[0109] (3) Closed-loop control reduces the risk of equipment failure. Based on real-time monitoring of the equipment's operating status and generating a closed-loop control instruction sequence, the equipment's operating parameters can be dynamically adjusted to respond to equipment abnormalities in real time, reducing the probability of equipment failure. Compared with traditional manual intervention or fixed rule control methods, this invention significantly improves the continuity and reliability of production.

[0110] Embodiment 2: Figure 2 FIG. 2 shows a block diagram of a vacuum isothermal forging press intelligent control system according to an embodiment of the present invention. Figure 2 As shown, the system may include: The data acquisition module 10 collects temperature, pressure and stress data during the forging process from a multi-modal sensor.

[0111] The sensors include thermocouples, infrared detectors, strain gauges and piezoelectric sensors, which acquire the temperature, pressure and stress data of the inner cavity of the press in real time.

[0112] The data is converted, amplified, filtered and other pre-processed by the signal processing unit and stored as a preliminary raw data sequence. ={T,P,σ}.

[0113] The data processing module 20 cleans and standardizes the collected data and generates high-quality working condition data.

[0114] Use Kalman filtering or wavelet denoising algorithm to denoise the data series. Perform noise filtering to remove outliers caused by sensor errors or environmental interference.

[0115] The z-score standardization method is used to calculate the mean μ and standard deviation σ of the data to obtain a standardized data sequence: .

[0116] The feature extraction and dynamic process prediction module 30 extracts key process features from standardized data and builds a dynamic process prediction model using a deep learning model.

[0117] Feature extraction: Fourier transform and principal component analysis methods are used to extract key features such as temperature change rate, pressure fluctuation amplitude and stress response.

[0118] Process prediction model: based on real-time operating data and historical data , using the recurrent neural network (RNN) model to predict future process parameters: .

[0119] The process parameter optimization module 40 optimizes the process parameters of the forging process using a reinforcement learning model.

[0120] The reinforcement learning model defines the state space S, action space A and reward function R. The state space S is the prediction parameter , the action space A is the temperature, pressure and speed adjustment ΔT, ΔP, Δv.

[0121] The process parameters are dynamically adjusted through the strategy π(S) and the optimal parameter set is output.

[0122] The closed-loop control and equipment management module 50 monitors the equipment operation status in real time and generates closed-loop control instructions.

[0123] Real-time monitoring: Collect the actual operating status of the equipment through sensors F(t)={T(t),P(t),v(t)}.

[0124] Instruction generation: based on feedback data F(t) and optimization parameters , adjust the closed-loop control instructions: .

[0125] Beneficial effects of the embodiments: Data processing and feature extraction: Multimodal sensors are used to collect and process high-precision operating data in real time, significantly improving the quality and reliability of data.

[0126] Dynamic prediction and optimization: Combine deep learning and reinforcement learning algorithms to dynamically adjust forging process parameters, effectively improving the quality consistency and production efficiency of forgings.

[0127] Closed-loop control and fault management: The system has real-time monitoring and closed-loop control functions, which can detect and repair faults in time, reduce equipment downtime and extend equipment life.

[0128] Full-process intelligent management: From data collection to process parameter optimization, and then to equipment management, full-process intelligent control of the forging process is achieved, which meets the intelligent manufacturing requirements of Industry 4.0.

[0129] Embodiment three: Figure 3 : is a block diagram of the structure of the intelligent control device for the vacuum isothermal forging press provided by the embodiment of the present invention. Figure 3 As shown, the device comprises: The data acquisition module (module 101) is used to collect multimodal data in the forging process in real time, including key process parameters such as temperature, pressure and stress.

[0130] The temperature T, pressure P and stress σ data in the forging environment are obtained through multimodal sensors such as thermocouples, strain gauges and infrared detectors.

[0131] The data is converted, amplified and filtered by the signal processing unit to generate a preliminary raw data sequence. .

[0132] The data processing and feature extraction module (module 102) performs noise filtering, outlier removal and standardization on the collected data, and extracts key process features.

[0133] Kalman filtering is used to remove noise, and statistical methods are combined to eliminate outliers.

[0134] Standardize the cleaned data to generate standardized data .

[0135] Extract key features (such as temperature change rate, pressure fluctuation amplitude, stress response) and generate feature data sets .

[0136] The dynamic process prediction module (module 103) uses deep learning models and expert knowledge to build a dynamic process prediction model to predict key process parameters in the forging process.

[0137] Input standardized data and historical data .

[0138] A dynamic prediction model is built through a recurrent neural network (RNN) to output future process parameter sequences: .

[0139] The process parameter optimization module (module 104) optimizes key process parameters in the forging process based on a reinforcement learning algorithm.

[0140] Enter prediction parameters As the state space, the action space is defined as the adjustment amount of temperature, pressure and press speed.

[0141] Use the reinforcement learning strategy π(S) to dynamically optimize the process parameters and output the optimal process parameter set.

[0142] The closed-loop control module (module 105) controls the press in real time according to the optimized process parameter set and generates closed-loop control instructions.

[0143] Monitor the equipment operating status F(t)={T(t),P(t),v(t)} and feedback the data.

[0144] Generate closed-loop control instruction sequence based on feedback data: .

[0145] Beneficial effects of the embodiments: Accurate data collection and processing. The present invention collects and processes forging data in real time through multimodal sensors, ensuring the accuracy and completeness of working condition data and providing high-quality data support for subsequent prediction and optimization.

[0146] Dynamic prediction and optimization. Combining deep learning models and reinforcement learning algorithms, the system can dynamically predict the trend of process parameter changes during the forging process and optimize process parameters in real time, improving the consistency and precision of forged products.

[0147] Intelligent closed-loop control. The system achieves real-time monitoring and dynamic adjustment through the closed-loop control module, ensuring that the equipment is always in the optimal operating state, effectively reducing human intervention.

[0148] Full process automation and intelligence. This device integrates data collection, feature extraction, parameter prediction, optimization adjustment and closed-loop control, realizing full process automation and intelligence of the forging process, greatly improving production efficiency and equipment utilization.

[0149] Obviously, the embodiments described above are only some embodiments of the present application, rather than all embodiments. The preferred embodiments of the present application are given in the accompanying drawings, but they do not limit the patent scope of the present application. The present application can be implemented in many different forms. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosure of the present application more thorough and comprehensive. Although the present application is described in detail with reference to the aforementioned embodiments, for those skilled in the art, it is still possible to modify the technical solutions recorded in the aforementioned specific implementation methods, or to perform equivalent replacement of some of the technical features therein. Any equivalent structure made using the contents of the specification and drawings of this application, directly or indirectly used in other related technical fields, is similarly within the scope of patent protection of this application.

Claims

1. A vacuum isothermal forging press intelligent control method, characterized in that: The method comprises: Based on multimodal sensors, the temperature, pressure and stress data of the forging process are collected to generate a preliminary raw data sequence; Cleaning and standardizing the original data sequence to generate standardized real-time operating condition data; Based on the real-time operating data and historical data, a deep learning model is used to extract key process features, a dynamic process prediction model is constructed, and a key parameter prediction sequence is generated; Inputting the key parameter prediction sequence into a reinforcement learning model, optimizing process parameters such as temperature, pressure and press speed during the forging process, and generating an optimized process parameter set; Based on the optimized process parameter set, the press execution system is dynamically adjusted, and the equipment operation status is monitored in real time to generate a closed-loop control instruction sequence.

2. The intelligent control method for vacuum isothermal forging press according to claim 1, characterized in that: The temperature, pressure and stress data collected during the forging process include: Acquire temperature data through thermocouples and infrared detectors; Obtain pressure data through strain gauges; The stress data is obtained through the mechanical sensor.

3. The intelligent control method for vacuum isothermal forging press according to claim 1, characterized in that: The cleaning and standardization of the original data sequence specifically includes: Using Kalman filtering algorithm to filter noise from the original data sequence; Detect and remove outliers to obtain cleaned data; The cleaned data is standardized based on the z-score standardization method to generate standardized real-time operating condition data.

4. The intelligent control method for vacuum isothermal forging press according to claim 1, characterized in that: The key process features of the extraction specifically include: Based on the standardized real-time operating data, Fourier transform is used to extract key features of temperature change rate, pressure fluctuation amplitude and stress response; The extracted features are screened using the principal component analysis method to generate a process feature data set.

5. The intelligent control method for vacuum isothermal forging press according to claim 1, characterized in that: The construction of the dynamic process prediction model specifically includes: Based on the process feature data set and historical process data, a dynamic process prediction model is established using a recurrent neural network model; The process feature data set is input, key parameters in the forging process are predicted, and a key parameter prediction sequence is generated.

6. The intelligent control method for vacuum isothermal forging press according to claim 1, characterized in that: The optimized forging process parameters specifically include: Using the key parameter prediction sequence as the state input of the reinforcement learning model; Optimize the process parameter adjustment strategy through reinforcement learning algorithm to generate the optimal process parameter combination; Output optimized temperature, pressure and press speed parameters.

7. The intelligent control method for vacuum isothermal forging press according to claim 1, characterized in that: The closed-loop control instruction generation specifically includes: Collect real-time operating status data of the equipment, including real-time temperature, pressure and press speed; generating a closed-loop control instruction sequence based on the feedback data and the optimized process parameter set; The temperature, pressure and press speed are continuously adjusted during the forging process.

8. The intelligent control method for vacuum isothermal forging press according to claim 1, characterized in that: Also includes: Use time series prediction models to analyze equipment operating status and generate fault warning information; Based on the fault warning information, the process parameters are dynamically corrected.

9. An intelligent control system for a vacuum isothermal forging press, characterized in that: The system comprises: A data acquisition module is used to collect temperature, pressure and stress data during the forging process from multimodal sensors to generate a raw data sequence; A data processing module, used for cleaning and standardizing the raw data sequence to generate standardized real-time operating condition data; A dynamic process prediction module is used to extract key process characteristics, build a dynamic process prediction model, and generate a key parameter prediction sequence based on the real-time operating data and historical data; A process parameter optimization module, used to optimize the process parameters based on the key parameter prediction sequence using a reinforcement learning model to generate an optimized process parameter set; The closed-loop control module is used to generate a closed-loop control instruction sequence based on the optimized process parameter set to dynamically adjust the press execution system.

10. An intelligent control device for a vacuum isothermal forging press, characterized in that: The device comprises a data acquisition module, a data processing and feature extraction module, a dynamic process prediction module, a process parameter optimization module, and a closed-loop control module. The device can implement the steps of the vacuum isothermal forging press intelligent control method as claimed in any one of claims 1 to 8.

Images