Intelligent control system and method for forging press clutch

Abstract

The application discloses a kind of forging equipment clutch intelligent regulation and control system and method, the system includes: including current sensor group, for real-time acquisition three-phase current waveform data of clutch drive motor;Feature extraction module is extracted waveform feature from waveform data, waveform feature includes harmonic distortion rate, phase shift and instantaneous power fluctuation characteristics;Timing prediction module is used to predict the future trend of clutch engagement state based on historical waveform characteristics;Controller is used to generate regulation and control instruction according to the prediction result, regulation and control instruction includes at least one of engagement time adjustment, motor torque compensation, lubrication control;According to the prediction result and feature parameter correlation analysis, dynamically generate engagement time adjustment, motor torque compensation, lubrication control and the like instruction, real-time optimization clutch operating state, reduce equipment abnormal vibration, meet the real-time monitoring and dynamic regulation and control demand of forging equipment clutch.

Description

Technical Field

[0001] This application relates to the field of forging equipment control technology, and in particular to an intelligent control system and method for a forging equipment clutch. Background Technology

[0002] In modern manufacturing, forging equipment is a key processing tool widely used in aerospace, automobile manufacturing, and machining. As the core component for power transmission in forging equipment, the clutch's control precision directly affects production efficiency and equipment stability.

[0003] Traditional clutch control methods often employ threshold-based switching control, relying on preset empirical parameters to achieve clutch engagement and disengagement. However, as the manufacturing industry moves towards intelligent and high-precision operations, this approach lacks adaptability to complex operating conditions and struggles to meet the demands of real-time monitoring and dynamic control. Summary of the Invention

[0004] The main technical problem addressed in this application is to provide an intelligent control system and method for the clutch of forging equipment, which meets the needs of real-time monitoring and dynamic control of the clutch of forging equipment.

[0005] To address the aforementioned technical problems, this application provides a technical solution: an intelligent control system for a forging press clutch. This system includes a current sensor group, a feature extraction module, a timing prediction module, and a controller. The current sensor group is used to collect waveform data of the three-phase current of the clutch drive motor in real time. The feature extraction module is used to extract waveform features from the waveform data, including harmonic distortion rate, phase offset, and instantaneous power fluctuation characteristics. The timing prediction module is used to predict the future trend of clutch engagement state based on historical waveform features. The controller is used to generate control commands based on the prediction results, including at least one of engagement time adjustment, motor torque compensation, and lubrication control. The current sensor group, feature extraction module, timing prediction module, and controller are connected via an industrial bus communication system.

[0006] Preferably, the generation of control commands is based on the correlation analysis of harmonic distortion rate, phase offset and instantaneous power fluctuation characteristics.

[0007] Preferably, the feature extraction module is configured to: decompose the waveform data into fundamental and harmonic components, calculate the total harmonic distortion rate; extract instantaneous power fluctuation features through time-frequency analysis, and identify abnormal power peaks at the moment of connection.

[0008] Preferably, the timing prediction module includes: a sliding time window unit for dividing continuous current waveform features into input segments according to the time sequence; and a prediction unit for predicting the probability distribution of the engagement state in a future preset time period based on the timing relationship of the input segments.

[0009] Preferably, the controller is configured to: generate a control command to reduce the output torque of the motor in response to the instantaneous power fluctuation continuously exceeding the preset steady-state range, and simultaneously start the lubrication system; generate a control command to shorten the engagement time in response to the harmonic distortion rate increasing higher than a first preset threshold and the phase offset exceeding a second preset threshold; and generate an emergency braking control command in response to the predicted clutch engagement failure probability exceeding a third preset threshold.

[0010] To address the aforementioned technical problems, another technical solution adopted in this application is: providing an intelligent control method for the clutch of a forging equipment, applied to any of the aforementioned intelligent control systems for the clutch of a forging equipment. This method includes: real-time acquisition of waveform data of the three-phase current of the clutch drive motor; extraction of waveform features from the waveform data, including harmonic distortion rate, phase offset, and instantaneous power fluctuation characteristics; prediction of future trends in clutch engagement state based on historical waveform features; and generation of control commands based on the prediction results, including at least one of engagement time adjustment, motor torque compensation, and lubrication control.

[0011] Preferably, the generation of control commands is based on the correlation analysis of harmonic distortion rate, phase offset and instantaneous power fluctuation characteristics.

[0012] Preferably, the method further includes: decomposing the waveform data into fundamental and harmonic components, calculating the total harmonic distortion rate; extracting instantaneous power fluctuation characteristics through time-frequency analysis, and identifying abnormal power peaks at the moment of connection.

[0013] Preferably, the method further includes: segmenting the continuous current waveform features into input segments according to the time sequence; and predicting the probability distribution of the engagement state in the future preset time period based on the temporal relationship of the input segments.

[0014] Preferably, the method further includes: in response to the instantaneous power fluctuation continuously exceeding the preset steady-state range, generating a control command to reduce the output torque of the motor and simultaneously starting the lubrication system; in response to the harmonic distortion rate increasing higher than a first preset threshold and the phase offset exceeding a second preset threshold, generating a control command to shorten the engagement time; and in response to the predicted clutch engagement failure probability exceeding a third preset threshold, generating an emergency braking control command.

[0015] The beneficial effects of this application are as follows: Unlike existing technologies, this application uses a current sensor group to collect three-phase current waveform data of the clutch drive motor in real time, and combines it with a feature extraction module to extract harmonic distortion rate, phase offset, and instantaneous power fluctuation characteristics, accurately reflecting the clutch engagement state and overcoming the limitations of traditional single-parameter monitoring; the timing prediction module segments historical waveform features based on a sliding time window, analyzes timing relationships through a prediction unit, and predicts the risk of clutch engagement failure in advance, realizing the transformation of fault warning from "post-event response" to "pre-event prevention"; the controller dynamically generates commands such as engagement time adjustment, motor torque compensation, and lubrication control based on the prediction results and feature parameter correlation analysis, optimizing the clutch operating state in real time, avoiding the decrease in forging accuracy caused by engagement time deviation, reducing abnormal equipment vibration, and meeting the real-time monitoring and dynamic control requirements of forging equipment clutches; furthermore, predictive maintenance strategies such as wear trend prediction are used to plan maintenance cycles in advance, reducing unplanned downtime and lowering equipment maintenance costs; at the same time, the emergency braking command triggering mechanism can avoid equipment damage caused by sudden failures and extend the service life of the clutch. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the framework of an embodiment of the intelligent control system for the clutch of the forging equipment of this application.

[0017] Figure 2 This is a flowchart illustrating an embodiment of the intelligent control method for the clutch of a forging press according to this application. Detailed Implementation

[0018] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application. The terms "comprising" and "having" and any variations thereof in this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0019] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0020] See Figure 1 , Figure 1 This is a schematic diagram of the framework of an embodiment of the intelligent control system for the clutch of the forging equipment of this application.

[0021] In this embodiment, the intelligent control system 10 for the clutch of the forging equipment includes a current sensor group 11, a feature extraction module 12, a timing prediction module 13, and a controller 14. The current sensor group 11 is used to collect waveform data of the three-phase current of the clutch drive motor in real time. The feature extraction module 12 is used to extract waveform features from the waveform data, including harmonic distortion rate, phase offset, and instantaneous power fluctuation characteristics. The timing prediction module 13 is used to predict the future trend of clutch engagement state based on historical waveform features. The controller 14 is used to generate control commands based on the prediction results, including at least one of engagement time adjustment, motor torque compensation, and lubrication control. The current sensor group 11, feature extraction module 12, timing prediction module 13, and controller 14 are connected via an industrial bus.

[0022] Specifically, the current sensor group 11 can employ a high-precision Hall effect sensor to acquire the time-domain waveform data of the three-phase current of the clutch drive motor in real time, thereby capturing the current spike at the moment of engagement and the harmonic components during steady-state operation; the feature extraction module 12 can integrate a fast Fourier transform algorithm unit to decompose the waveform data into the fundamental wave and each harmonic component. By calculating the square root of the sum of the squares of the effective values ​​of each harmonic current, dividing it by the effective value of the fundamental current, and multiplying by 100%, the total harmonic distortion rate is obtained. At the same time, wavelet transform is used for time-frequency analysis to extract the instantaneous power fluctuation curve with millisecond-level resolution to identify abnormal power peaks during the engagement impact stage, such as transient spikes exceeding 150% of the steady-state value; the sliding time window unit of the timing prediction module 13 can be set to be adjustable from 10 seconds to 30 seconds, dividing the continuous feature data sequence into input segments of equal duration. The prediction unit can be based on a long short-term memory network or a hidden Markov model, and through training historical data, including samples of normal engagement, half-engagement faults, and complete failures, establish a state transition probability matrix to realize the future 1 Within minutes, the engagement status includes the probability distribution prediction of normal, warning, and failure. The controller 14 can be connected to each module via industrial buses such as PROFIBUS-DP or Ethernet / IP, and uses communication protocols such as Modbus TCP to achieve real-time data interaction. Its built-in decision algorithm library can include rule-based fuzzy logic controllers and model predictive control algorithms. It can dynamically generate pulse width modulation signals to adjust motor drive parameters or output digital signals to control the opening of solenoid valves in the lubrication system based on the multi-parameter coupling analysis results, such as the increase in harmonic distortion rate (e.g., more than 5% per second) and phase offset (e.g., the fundamental phase difference exceeds 15°).

[0023] The above solution is successful. The current sensor group 11 collects the three-phase current waveform data of the clutch drive motor in real time. Combined with the feature extraction module 12, it extracts harmonic distortion rate, phase offset, and instantaneous power fluctuation characteristics, accurately reflecting the clutch engagement state and overcoming the limitations of traditional single-parameter monitoring. The timing prediction module 13 segments historical waveform features based on a sliding time window and analyzes timing relationships through a prediction unit to predict clutch engagement failure risks in advance, transforming fault warning from "post-event response" to "pre-event prevention." The controller 14 dynamically generates commands such as engagement time adjustment, motor torque compensation, and lubrication control based on the prediction results and feature parameter correlation analysis, optimizing the clutch operating state in real time, avoiding forging accuracy degradation caused by engagement time deviations, reducing abnormal equipment vibration, and meeting the real-time monitoring and dynamic control requirements of the forging equipment clutch. Furthermore, predictive maintenance strategies, such as wear trend prediction, allow for advance planning of maintenance cycles, reducing unplanned downtime and lowering equipment maintenance costs. Simultaneously, the emergency braking command triggering mechanism can prevent equipment damage caused by sudden failures and extend clutch lifespan.

[0024] In some embodiments, the generation of control commands is based on the correlation analysis of harmonic distortion rate, phase offset, and instantaneous power fluctuation characteristics.

[0025] Specifically, the controller 14 can establish a mapping relationship between feature parameters and clutch state through a multi-parameter fusion algorithm. For example, it can set weight coefficients for different features and construct a decision matrix: harmonic distortion rate reflects the degree of distortion of the motor current waveform, phase offset reflects the change in voltage and current phase difference, and instantaneous power fluctuation indicates the energy impact during engagement. The controller 14 can comprehensively judge the coupling state of the three through preset logic rules or machine learning models. For example, if the harmonic distortion rate exceeds 20% of the steady-state value for three consecutive sampling periods and the phase offset is greater than 10°, and the peak-to-peak value of the instantaneous power fluctuation exceeds 120% of the rated power, it is determined to be a "engagement impact risk state", thereby triggering a targeted control strategy.

[0026] In some specific embodiments, the correlation analysis may include time-series correlation analysis and threshold cross-validation: First, the harmonic distortion rate growth rate (i.e., the change in distortion rate per unit time, the absolute value of phase offset, and the instantaneous power fluctuation amplitude) is normalized. Then, the Pearson correlation coefficient of the three is calculated through a sliding window. When the correlation coefficient exceeds 0.7, it is determined to be a strong correlation. For example, if the harmonic distortion rate growth rate is higher than 5% per second and the phase offset exceeds 15° (the second preset threshold), and the instantaneous power fluctuation lasts for 500ms and exceeds the steady-state range of ±15%, the controller 14 will generate a command to shorten the engagement time by 10%-15% to reduce the wear of the clutch disc by the impact load.

[0027] In some application scenarios, when forging equipment performs high-precision forging, such as forming aerospace parts, if the instantaneous power fluctuation is detected to have a peak exceeding 180% of the steady-state value at the moment of engagement, and the harmonic distortion rate increases by 8% per second, while the phase offset exceeds 20°, the system will determine it to be a "critical failure state". At this time, the controller 14 will prioritize generating an emergency braking command to cut off the motor power and activate the mechanical braking device to avoid damage to the forging die due to clutch failure. Under normal operating conditions, if only the instantaneous power fluctuation is detected to exceed the steady-state range but the multi-parameter linkage threshold is not triggered, the controller 14 will only generate a motor torque compensation command to ensure smooth engagement by increasing the drive torque by 5%-10%, while recording abnormal data for subsequent iterative optimization of the wear prediction model.

[0028] In some embodiments, the feature extraction module 12 is configured to: decompose the waveform data into fundamental and harmonic components, calculate the total harmonic distortion rate; extract instantaneous power fluctuation features through time-frequency analysis, and identify abnormal power peaks at the moment of connection.

[0029] Specifically, the feature extraction module 12 can integrate a signal processing unit. First, it preprocesses the time-domain waveform data acquired by the current sensor group 11, filtering out high-frequency noise using a digital filter. Then, it uses a fast Fourier transform to convert the time-domain signal into a frequency-domain signal, separating the 50Hz fundamental component and second-order and higher harmonic components. The total harmonic distortion rate is obtained by calculating the square root of the sum of the squares of the effective values ​​of each harmonic current, dividing it by the effective value of the fundamental current, and multiplying by 100%, thus quantifying the distortion degree of the current waveform. Simultaneously, wavelet transform, such as the db4 wavelet basis, is used for time-frequency analysis to generate a three-dimensional time-frequency-energy spectrum. Instantaneous power curves with millisecond-level resolution are extracted from this spectrum. By setting a dynamic threshold, such as 130% of the steady-state power, abnormal power peaks at the moment of connection are identified.

[0030] In some specific embodiments, the signal processing unit can employ a field-programmable gate array (FPGA) or a digital signal processor (DSP) to achieve high-speed computation and support real-time online processing. For example, for the impact current during the start-up phase of the forging equipment, the feature extraction module 12 can automatically switch to a high-frequency sampling mode, such as 20kHz, to ensure the capture of the current surge at the moment of contact; for steady-state operating data, the sampling frequency can be reduced to 10kHz to reduce the computational load. Furthermore, the algorithm for identifying abnormal power peaks can be combined with morphological filtering to remove narrowband interference signals and avoid misjudgments.

[0031] In some application scenarios, when forging equipment undergoes multiple die switching or sudden load changes, the feature extraction module 12 can monitor in real time the step change in the amplitude of the fundamental component and the significant increase in harmonic components. For example, in the clutch semi-engaged state, the instantaneous power fluctuation curve will show periodic spikes. Since the frequency is related to the clutch speed, the module can accurately identify the semi-engaged fault by comparing the feature templates of normal engagement and abnormal states in the historical database, providing an early warning basis for the controller 14. During the equipment preheating stage, the module can determine whether the lubrication system is working properly by analyzing the changing trend of the fundamental phase offset, and indirectly assess whether the friction coefficient of the clutch plate is within a reasonable range.

[0032] In some embodiments, the timing prediction module 13 includes: a sliding time window unit for dividing continuous current waveform features into input segments according to a time sequence; and a prediction unit for predicting the probability distribution of the engagement state in a future preset time period based on the timing relationship of the input segments.

[0033] Specifically, the sliding time window unit supports dynamic adjustment of the window length and overlap rate. For example, the window length can be set to 10 to 60 seconds, and the overlap rate to 20%-50%, to balance the correlation of historical data with real-time response speed. The input segment contains characteristic parameters such as harmonic distortion rate, phase shift, and instantaneous power fluctuation of the continuous time series, forming a multi-dimensional time series data matrix, such as time step × feature dimension. The prediction unit can use algorithms such as long short-term memory networks, time series models, or gradient boosting trees. By training on historical data covering multiple operating conditions such as normal operation, wear and aging, and lubrication failure, it learns the time dependence of characteristic parameters and state transition rules, and outputs the clutch engagement status within the next 1 to 10 minutes, including the probability distribution of normal, warning, and failure.

[0034] In some specific embodiments, the sliding time window unit can automatically switch window strategies according to the equipment operating mode: when the forging equipment is in high-frequency, high-speed forging mode, a 10-second window length and a 50% overlap rate are used to ensure real-time capture of transient feature changes; when in low-speed, heavy-load mode, it switches to a 30-second window length and a 20% overlap rate to focus on long-term trend analysis. The prediction unit can integrate an attention mechanism to assign higher weight to recent features, improving prediction sensitivity under abrupt changes. For example, for abnormal power peak sequences in the input segment, the attention mechanism can strengthen the influence weight of this feature on future states, thereby improving prediction accuracy.

[0035] In some application scenarios, when forging equipment enters the fatigue stage after continuous operation for more than 8 hours, the timing prediction module 13 can predict that the probability of clutch plate wear leading to engagement failure exceeds 60% within the next 2 hours by analyzing the gradual upward trend of harmonic distortion rate and the periodic fluctuation of phase offset in historical data. This triggers a preventive maintenance reminder, prompting operators to check the lubrication system or replace vulnerable parts. In smart factory scenarios where multiple devices are networked, the prediction unit can upload the engagement state probability distribution of a single device to the cloud. By training and optimizing the global prediction model through cluster data, it can predict common fault trends of devices of the same model, helping factories plan downtime maintenance cycles in advance and reduce the risk of unplanned production line interruptions.

[0036] In some embodiments, the controller 14 is configured to: generate a control command to reduce the output torque of the motor in response to the instantaneous power fluctuation continuously exceeding a preset steady-state range, and simultaneously start the lubrication system; generate a control command to shorten the engagement time in response to the harmonic distortion rate increasing above a first preset threshold and the phase offset exceeding a second preset threshold; and generate an emergency braking control command in response to the predicted clutch engagement failure probability exceeding a third preset threshold.

[0037] Specifically, from the perspective of instantaneous power fluctuations, the preset steady-state range is determined based on a large amount of experimental data and the characteristics of the equipment during normal operation. For example, for a certain model of forging equipment, after multiple tests, it was found that during stable operation, the instantaneous power fluctuation within ±10% of the rated power is normal. Therefore, this range is set as the preset steady-state range. When the instantaneous power fluctuations monitored in real time continuously exceed this range, it means that the equipment may be in an overload or abnormal operating state. At this time, a control command to reduce the motor output torque is generated to reduce the load on the equipment. The controller 14 sends an adjustment signal to the motor driver to change parameters such as the motor's input current or voltage, thereby reducing the motor's output torque. Generally, the motor output torque will be reduced by 10% - 20%, a percentage determined by comprehensively considering the equipment's capacity and the possibility of restoring normal operation. The lubrication system is started simultaneously because under abnormal operating conditions, the friction between the clutch plates will increase, easily leading to overheating and wear. After the lubrication system is started, the injection of grease is controlled by the solenoid valve. The grease is injected at the maximum opening for 2-5 seconds, which can effectively reduce friction and protect the clutch disc.

[0038] From the perspective of harmonic distortion rate growth and phase offset, harmonic distortion rate reflects the degree of distortion in the current waveform, while its growth rate reflects how this distortion changes over time. The first preset threshold is set based on the equipment's tolerance to harmonic distortion, for example, 5% per second. Phase offset refers to the phase difference between voltage and current. Under normal circumstances, the phases of voltage and current should remain relatively stable. When it exceeds a certain range, it indicates that the equipment is malfunctioning. The second preset threshold is set, for example, to 15° electrical angle. When the harmonic distortion rate growth rate exceeds the first preset threshold and the phase offset exceeds the second preset threshold, it indicates that the clutch's operating state has deviated from normal, and problems such as unstable engagement may occur. At this time, a control command to shorten the engagement time is generated. The controller 14 adjusts the delayed trigger signal of the drive circuit to change the clutch engagement time, shortening the clutch engagement time by 10%-15%, thereby reducing asynchronous impact and ensuring the normal operation of the clutch.

[0039] From the perspective of clutch engagement failure probability, the third preset threshold is determined based on the assessment of equipment failure risk and acceptable safety level, for example, set to 70%. If the predicted clutch engagement failure probability exceeds this threshold, it indicates a high likelihood of equipment failure. If measures are not taken promptly, it could lead to serious equipment damage and production accidents. Therefore, controller 14 generates an emergency braking control command. First, it sends a power-off command to the motor driver, cutting off the motor's power input and stopping the motor. Simultaneously, it sends a trigger signal to the mechanical braking mechanism, causing the mechanical braking device to act rapidly, forcibly disengaging the clutch within 50ms to prevent further escalation of the fault and ensure the safety of equipment and personnel.

[0040] In some specific embodiments, the generation strategy of the control commands can support customized parameter configuration, allowing users to adjust preset thresholds and execution parameters according to equipment model and processing technology through a human-machine interface. For example, for large-tonnage forging equipment, the steady-state range can be extended to ±15% of the rated power to reduce the frequency of false triggering; for high-speed precision equipment, the reduction ratio of engagement time is limited to 5%-10% to avoid excessive adjustment affecting processing accuracy. The controller 14 hardware can adopt an industrial-grade programmable logic controller or an embedded microprocessor, integrating a real-time operating system to ensure millisecond-level response of control commands, and communicating with the host computer monitoring system via Modbus / TCP protocol to upload control logs and abnormal event records in real time.

[0041] In some application scenarios, when the forging equipment processes high-hardness alloy materials, causing a sudden increase in load, the instantaneous power fluctuation may exceed the steady-state range by 30% at the moment of engagement and last for more than 1 second. The controller 14 immediately reduces the motor torque by 15% and activates the lubrication system to prevent the clutch plates from burning due to overload friction. If bearing wear occurs after prolonged operation, causing the harmonic distortion rate to increase by 8% per second and the phase offset to exceed 20°, the controller 14 automatically shortens the engagement time by 12% to compensate for the response delay caused by the aging of mechanical components. When the prediction unit outputs an engagement failure probability exceeding 75% for three consecutive times, the controller 14 generates an emergency braking command and sends a fault code to the factory's IoT platform, triggering a three-level early warning mechanism to prompt maintenance personnel to prioritize the repair of the equipment and avoid the risk of a chain reaction of production line shutdowns. Through multi-strategy collaborative control, the response time of abnormal clutch conditions can be controlled within 200ms, effectively improving the safety and reliability of equipment operation.

[0042] The above solution uses a current sensor group 11 to collect three-phase current waveform data of the clutch drive motor in real time. Combined with the feature extraction module 12, it extracts harmonic distortion rate, phase offset, and instantaneous power fluctuation characteristics to accurately reflect the clutch engagement state, overcoming the limitations of traditional single-parameter monitoring. The timing prediction module 13 segments historical waveform features based on a sliding time window and analyzes timing relationships through a prediction unit to predict the risk of clutch engagement failure in advance, realizing the transformation of fault warning from "post-event response" to "pre-event prevention". The controller 14 dynamically generates commands such as engagement time adjustment, motor torque compensation, and lubrication control based on the prediction results and feature parameter correlation analysis, optimizing the clutch operating state in real time, avoiding the decrease in forging accuracy caused by engagement time deviation, reducing abnormal equipment vibration, and meeting the real-time monitoring and dynamic control requirements of the forging equipment clutch. Furthermore, predictive maintenance strategies such as wear trend prediction are used to plan maintenance cycles in advance, reducing unplanned downtime and lowering equipment maintenance costs. At the same time, the emergency braking command triggering mechanism can avoid equipment damage caused by sudden failures and extend the service life of the clutch.

[0043] Please see Figure 2 , Figure 2 This is a flowchart illustrating an embodiment of the intelligent control method for the clutch of a forging press according to this application. In this embodiment, the method includes the following steps: S21: Real-time acquisition of waveform data of the three-phase current of the clutch drive motor.

[0044] S22: Extract waveform features from waveform data, including harmonic distortion rate, phase offset, and instantaneous power fluctuation characteristics.

[0045] S23: Predict future trends in clutch engagement state based on historical waveform characteristics.

[0046] S24: Generate control instructions based on the prediction results. The control instructions include at least one of engagement time adjustment, motor torque compensation, and lubrication control.

[0047] In some embodiments, the generation of control commands is based on the correlation analysis of harmonic distortion rate, phase offset, and instantaneous power fluctuation characteristics.

[0048] In some embodiments, the feature extraction module is configured to: decompose the waveform data into fundamental and harmonic components, calculate the total harmonic distortion rate; extract instantaneous power fluctuation features through time-frequency analysis, and identify abnormal power peaks at the moment of connection.

[0049] In some embodiments, the method further includes: segmenting continuous current waveform features into input segments according to a time series; and predicting the probability distribution of engagement states in a future preset time period based on the temporal relationship of the input segments.

[0050] In some embodiments, the method further includes: generating a control command to reduce the output torque of the motor in response to the instantaneous power fluctuation continuously exceeding a preset steady-state range, and simultaneously starting the lubrication system; generating a control command to shorten the engagement time in response to the harmonic distortion rate increasing above a first preset threshold and the phase offset exceeding a second preset threshold; and generating an emergency braking control command in response to the predicted clutch engagement failure probability exceeding a third preset threshold.

[0051] The above solution accurately reflects the clutch engagement state by real-time acquisition of three-phase current waveform data of the clutch drive motor, combined with the feature extraction module to extract harmonic distortion rate, phase offset, and instantaneous power fluctuation characteristics, thus overcoming the limitations of traditional single-parameter monitoring. Based on the sliding time window segmentation of historical waveform features, the solution analyzes the timing relationship to predict the risk of clutch engagement failure in advance, realizing the transformation of fault early warning from "post-event response" to "pre-event prevention." Based on the prediction results and feature parameter correlation analysis, the solution dynamically generates commands such as engagement time adjustment, motor torque compensation, and lubrication control, optimizing the clutch operating state in real time, avoiding the decrease in forging accuracy caused by engagement time deviation, reducing abnormal equipment vibration, and meeting the real-time monitoring and dynamic control requirements of forging equipment clutches. Furthermore, predictive maintenance strategies, such as wear trend prediction, allow for advance planning of maintenance cycles, reducing unplanned downtime and lowering equipment maintenance costs. At the same time, the emergency braking command triggering mechanism can prevent equipment damage caused by sudden failures and extend the service life of the clutch.

[0052] The above description is merely an embodiment of this application and does not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. An intelligent control system for the clutch of a forging press, characterized in that, include: A current sensor array is used to acquire waveform data of the three-phase current of the clutch drive motor in real time. The feature extraction module is used to extract waveform features from the waveform data, including harmonic distortion rate, phase offset and instantaneous power fluctuation characteristics; The timing prediction module is used to predict future trends in clutch engagement state based on historical waveform characteristics. The controller is configured to: In response to the instantaneous power fluctuation continuously exceeding the preset steady-state range, a control command to reduce the motor output torque is generated, and the lubrication system is started simultaneously; and / or; In response to the harmonic distortion rate growth rate being higher than a first preset threshold and the phase offset exceeding a second preset threshold, a control command to shorten the engagement time is generated. and / or; In response to the prediction that the probability of clutch engagement failure exceeds a third preset threshold, an emergency braking control command is generated. The current sensor group, the feature extraction module, the timing prediction module, and the controller are connected via an industrial bus.

2. The intelligent control system for the clutch of forging equipment according to claim 1, characterized in that, The generation of the control command is based on the correlation analysis of the harmonic distortion rate, phase offset, and instantaneous power fluctuation characteristics.

3. The intelligent control system for the clutch of forging equipment according to claim 1, characterized in that, The feature extraction module is configured as follows: The waveform data is decomposed into fundamental and harmonic components, and the total harmonic distortion rate is calculated. Instantaneous power fluctuation characteristics are extracted through time-frequency analysis to identify abnormal power peaks at the moment of connection.

4. The intelligent control system for the clutch of forging equipment according to claim 3, characterized in that, The time series prediction module includes: Sliding time window unit: used to divide continuous current waveform features into input segments according to the time series; Prediction unit: used to predict the probability distribution of the joining state in a future preset time period based on the temporal relationship of the input segment.

5. A method for intelligent control of the clutch in a forging press, characterized in that, The method, applied to the intelligent control system for the clutch of a forging equipment as described in any one of claims 1-4, comprises: Real-time acquisition of waveform data of the three-phase current of the clutch drive motor; Waveform features are extracted from the waveform data, including harmonic distortion rate, phase offset, and instantaneous power fluctuation characteristics; Predicting future trends in clutch engagement state based on historical waveform characteristics; The method further includes: In response to the instantaneous power fluctuation continuously exceeding the preset steady-state range, a control command to reduce the motor output torque is generated, and the lubrication system is started simultaneously; and / or; In response to the harmonic distortion rate growth rate being higher than a first preset threshold and the phase offset exceeding a second preset threshold, a control command to shorten the engagement time is generated. and / or; In response to the prediction that the probability of clutch engagement failure exceeds a third preset threshold, an emergency braking control command is generated.

6. The intelligent control method for the clutch of a forging press according to claim 5, characterized in that, The generation of the control command is based on the correlation analysis of the harmonic distortion rate, phase offset, and instantaneous power fluctuation characteristics.

7. The intelligent control method for the clutch of forging equipment according to claim 5, characterized in that, The method further includes: The waveform data is decomposed into fundamental and harmonic components, and the total harmonic distortion rate is calculated. Instantaneous power fluctuation characteristics are extracted through time-frequency analysis to identify abnormal power peaks at the moment of connection.

8. The intelligent control method for the clutch of a forging press according to claim 7, characterized in that, The method further includes: segmenting continuous current waveform features into input segments according to a time series; Based on the temporal relationship of the input segments, predict the probability distribution of the junction state in the future preset time period.

Images