Mahjong tile breaking machine feeding dynamic control method and system

Abstract

The invention belongs to the technical field of industrial equipment control, and discloses a dynamic control method and system for feeding of a card breaking machine, and the method comprises the steps: obtaining the multi-dimensional state characteristics of the card breaking machine, and generating a recommended feeding control action through an automatic feeding control strategy generation logic; monitoring an actual feeding control action executed by the operator through the tile breaking machine, and comparing the actual feeding control action with the recommended feeding control action to identify a correction behavior of the operator; when the correction behavior of the operator is identified, monitoring equipment operation data of the mahjong machine in an observation time period after the actual feeding control action is executed, so as to evaluate multi-dimensional equipment performance improvement information; according to the multi-dimensional state features, the recommended feeding control action, the actual feeding control action and the multi-dimensional equipment performance improvement information, an automatic feeding control strategy generation logic is adjusted so as to perform feeding control on the tile breaking machine; therefore, the experience of an operator can be learned, and the self-adaptive capability and performance of the tile breaking machine under complex working conditions are improved.

Description

Technical Field

[0001] The present application relates to the field of industrial equipment control technology, and in particular to a dynamic control method and system for feeding a card crusher. Background Art

[0002] In waste plastic recycling lines, plastic flake shredders are key equipment responsible for reducing waste plastic flakes to the desired size. Their performance and safe operation are closely linked to the control of the feeder. Ideally, feeder control should dynamically adjust based on material characteristics and equipment status, ensuring a balanced balance between efficiency and safety.

[0003] Existing plastic shredders typically utilize automated feed control systems based on key parameters such as motor current and spindle speed. These systems rely on preset logic or simple feedback algorithms, such as reducing the speed when current exceeds a threshold. However, the characteristics of the waste plastic flakes actually processed are complex and varied, with significant variations in material, thickness, shape, size, moisture content, and impurities. This leads to complex variations in resistance and energy requirements during the shredding process. Automated systems based on simple parameter feedback struggle to accurately capture this complex material-equipment interaction, and are prone to control lag or ineffectiveness under complex operating conditions, making it difficult to continuously maintain an optimal or safe state.

[0004] In actual production, experienced operators can compensate for the shortcomings of automated systems by observing multi-dimensional equipment signals. For example, an operator might detect abnormalities in the equipment's sound spectrum or current waveform, predicting potential risks and initiating manual intervention even if critical parameters fall below thresholds. These intervention strategies, reflecting operator experience, enable more effective responses to complex situations.

[0005] However, existing automation systems are typically only able to record simple speed adjustments made by operators and are unable to identify and record more complex interventions, such as short pauses and mode switches, that operators perform based on multi-dimensional signal judgments. More importantly, existing systems are unable to systematically correlate and learn from these operator interventions with the multi-modal operating status data of the equipment at the time of the intervention, as well as the multi-dimensional improvements in the equipment's operating status (such as efficiency, energy consumption, and stability) after the intervention. This makes it difficult for automation systems to learn and replicate the operator's advanced experience strategies, limiting the performance ceiling and adaptive capabilities of the shredder when processing complex materials.

[0006] In view of the above problems, the existing technology is in urgent need of improvement. Summary of the Invention

[0007] The purpose of this application is to provide a dynamic control method and system for feeding a card shredder, which can learn from the operator's experience and improve the adaptability and performance of the card shredder under complex working conditions.

[0008] In a first aspect, the present application provides a dynamic control method for feeding a plastic chip shredder, which is used to control the feeding action of the plastic chip shredder. The method comprises the following steps:

[0009] A1. Obtain multi-dimensional device operation signals from the card shredder and extract multi-dimensional status features from them;

[0010] A2. Generate recommended feed control actions based on the multi-dimensional state characteristics using automated feed control strategy generation logic;

[0011] A3. Monitoring the actual feed control action performed by the operator to crush the card machine, comparing the actual feed control action with the recommended feed control action to identify the operator's corrective behavior;

[0012] A4. When an operator's corrective action is identified, monitor the card shredder equipment operating data during the observation period after the actual feed control action is executed to evaluate multi-dimensional equipment performance improvement information;

[0013] A5. Generate a corrective feedback signal based on the multi-dimensional state characteristics, the recommended feed control action, the actual feed control action, and the multi-dimensional device performance improvement information;

[0014] A6. Adjust the automated feed control strategy generation logic according to the corrective feedback signal;

[0015] A7. Generate logic based on the adjusted automated feeding control strategy to control the feeding of the card crusher.

[0016] Preferably, step A1 includes:

[0017] A101. Get the multi-dimensional device operation signal of the card crusher; the multi-dimensional device operation signal includes a motor current signal, a device sound signal, and a device vibration signal;

[0018] A102 performs waveform analysis on the motor current signal to extract impact load characteristics;

[0019] A103 performs spectrum analysis on the device sound signal to extract sound spectrum features;

[0020] A104. Perform spectrum analysis on the vibration signal of the device to extract vibration spectrum characteristics;

[0021] A105. Combine the impact load characteristics, the sound spectrum characteristics, and the vibration spectrum characteristics to form the multi-dimensional state characteristics.

[0022] Preferably, step A3 includes:

[0023] A301. Get the actual control instructions and parameter settings executed by the operator through the card crusher control interface;

[0024] A302. Analyze the actual control instructions and parameter settings to identify the type and parameters of the actual feed control action;

[0025] A303. Get the type and parameters of the recommended feed control action;

[0026] A304. Compare the type and parameters of the actual feed control action with the type and parameters of the recommended feed control action to determine whether there is a difference between the actual feed control action and the recommended feed control action;

[0027] A305. When there is a difference, it is determined that there is a corrective action, and the actual feed control action is used as the operator's corrective action.

[0028] Preferably, step A4 includes:

[0029] A401. When the operator's corrective action is identified, determining the observation period;

[0030] A402. During the observation period, monitor multi-dimensional equipment operating data for the card shredder; the multi-dimensional equipment operating data includes average motor power, motor current, equipment sound signal energy within a specific frequency range, equipment vibration signal amplitude within a specific frequency range, and the card shredder's discharge volume and discharge particle size distribution.

[0031] A403. Evaluate multi-dimensional equipment performance improvement information based on the multi-dimensional equipment operation data; the multi-dimensional equipment performance improvement information includes efficiency improvement indicators, energy consumption reduction indicators and stability enhancement indicators.

[0032] Preferably, step A401 includes:

[0033] Extracting state features reflecting characteristics of the currently processed material from the multi-dimensional state features as material characterization features;

[0034] The observation time period is determined according to the material characterization characteristics and the type of the actual feeding control action, according to a preset observation time period determination rule or an observation time period lookup table.

[0035] Preferably, step A403 includes:

[0036] B1. Calculate the evaluation values ​​of the efficiency improvement index, the energy consumption reduction index, and the stability enhancement index based on the multi-dimensional device operation data;

[0037] B2. Identify information reflecting the corrective behavior goal based on the actual feed control action, recorded as corrective intention information;

[0038] B3. Determine the weight coefficients of the efficiency improvement index, the energy consumption reduction index, and the stability enhancement index based on the material characterization characteristics and the correction intention information;

[0039] B4. Calculate the comprehensive performance improvement index based on the evaluation values ​​of the efficiency improvement index, the energy consumption reduction index, and the stability enhancement index and the corresponding weight coefficients;

[0040] B5. The evaluation values ​​of the efficiency improvement index, the energy consumption reduction index, the stability enhancement index, and the comprehensive performance improvement index are combined into the multi-dimensional equipment performance improvement information.

[0041] Preferably, step B2 includes:

[0042] Based on a preset intention recognition rule or an intention lookup table, the correction intention information is identified according to the type and parameters of the actual feed control action and the multi-dimensional state characteristics.

[0043] Preferably, step A6 includes:

[0044] A601. Based on the multi-dimensional state characteristics in the corrected feedback signal, locate the strategy parameters or rules corresponding to the multi-dimensional state characteristics in the automated feed control strategy generation logic as target strategy parameters or target rules;

[0045] A602. Calculate the feedback intensity of the actual feed control action based on the evaluation value of the efficiency improvement index, the evaluation value of the energy consumption reduction index, the evaluation value of the stability enhancement index, and the comprehensive performance improvement index in the multi-dimensional equipment performance improvement information in the corrected feedback signal;

[0046] A603. Adjust the target strategy parameters or target rules according to the feedback strength to increase the probability of generating the actual feed control action under a state similar to the multi-dimensional state characteristics.

[0047] Preferably, step A602 includes:

[0048] Determining preliminary feedback intensity based on the comprehensive performance improvement indicator;

[0049] determining intention compliance based on the correction intention information, the evaluation value of the efficiency improvement index, the evaluation value of the energy consumption reduction index, and the evaluation value of the stability enhancement index;

[0050] The preliminary feedback strength is modified according to the intention compliance to obtain a final feedback strength.

[0051] In a second aspect, the present application provides a dynamic control system for feeding a plastic chip shredder, which is used to control the feeding action of the plastic chip shredder. The system includes:

[0052] A state feature acquisition module is used to obtain the multi-dimensional device operation signal of the card shredder and extract multi-dimensional state features from it;

[0053] A recommended action generation module, configured to generate a recommended feed control action based on the multi-dimensional state characteristics using an automated feed control strategy generation logic;

[0054] a corrective action identification module, configured to monitor actual feed control actions performed by the operator on the card shredder, and compare the actual feed control actions with the recommended feed control actions to identify the operator's corrective actions;

[0055] a performance evaluation module for monitoring the card crusher equipment operating data during an observation period after the actual feed control action is executed when the operator's corrective action is identified, so as to evaluate multi-dimensional equipment performance improvement information;

[0056] a feedback signal generating module, configured to generate a corrective feedback signal based on the multi-dimensional state characteristics, the recommended feed control action, the actual feed control action, and the multi-dimensional equipment performance improvement information;

[0057] A strategy adjustment module, configured to adjust the automatic feeding control strategy generation logic according to the correction feedback signal;

[0058] The control execution module is used to generate logic according to the adjusted automatic feeding control strategy to control the feeding of the card crusher.

[0059] Beneficial effects: The present application provides a dynamic control method and system for feeding a card shredder, which solves the problem of difficulty in dealing with complex materials and learning operator experience in the prior art by acquiring multi-dimensional equipment signals, identifying the operator's corrective behavior and learning from his experience. It has the advantage of being able to learn from the operator's experience and improve the adaptability and performance of the card shredder under complex working conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 This is a flow chart of the dynamic control method for feeding a card crusher provided in an embodiment of the present application.

[0061] Figure 2 This is a structural diagram of the dynamic control system for feeding a card crusher provided in an embodiment of the present application.

[0062] Explanation of the numbers: 1. State feature acquisition module; 2. Recommended action generation module; 3. Corrective behavior recognition module; 4. Performance evaluation module; 5. Feedback signal generation module; 6. Strategy adjustment module; 7. Control execution module. DETAILED DESCRIPTION

[0063] The technical solutions in this application will be clearly and completely described below in conjunction with the drawings in this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all of the embodiments. The components of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application for protection, but merely represents selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work fall within the scope of protection of this application.

[0064] It should be noted that similar reference numerals and letters represent similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings. At the same time, in the description of this application, the terms "first", "second", etc. are only used to distinguish the description and should not be understood as indicating or implying relative importance.

[0065] refer to Figure 1 This application proposes a dynamic control method for feeding a plastic chip shredder, which is used to control the feeding action of the plastic chip shredder. The method comprises the following steps:

[0066] A1. Obtain multi-dimensional device operation signals from the card shredder and extract multi-dimensional status features from them;

[0067] A2. Generate recommended feed control actions based on the multi-dimensional state characteristics using automated feed control strategy generation logic;

[0068] A3. Monitoring the actual feed control action performed by the operator to crush the card machine, comparing the actual feed control action with the recommended feed control action to identify the operator's corrective behavior;

[0069] A4. When an operator's corrective action is identified, monitor the card shredder equipment operating data during the observation period after the actual feed control action is executed to evaluate multi-dimensional equipment performance improvement information;

[0070] A5. Generate a corrective feedback signal based on the multi-dimensional state characteristics, the recommended feed control action, the actual feed control action, and the multi-dimensional device performance improvement information;

[0071] A6. Adjust the automated feed control strategy generation logic according to the corrective feedback signal;

[0072] A7. Generate logic based on the adjusted automated feeding control strategy to control the feeding of the card crusher.

[0073] Among them, multi-dimensional equipment operation signals refer to various types of sensor data or monitoring data that reflect the current working status of the card shredder. They can be obtained using current sensors, sound sensors, vibration sensors, speed sensors, temperature sensors and other devices, such as motor current signals, equipment sound signals, equipment vibration signals, etc., mainly to fully perceive the real-time operating status of the card shredder and the characteristics of the processed materials.

[0074] Among them, multi-dimensional state features refer to key information extracted from multi-dimensional equipment operation signals that can characterize the status of equipment and materials. They can be implemented using signal processing and feature extraction technologies. For example, waveform analysis of motor current signals can be performed to extract impact load features, and spectrum analysis of equipment sound signals and vibration signals can be performed to extract spectrum features. The main purpose is to convert raw operating data into structured information that can be used for control decisions and status judgment.

[0075] Among them, the automated feeding control strategy generation logic refers to the algorithm or model used to automatically determine the feeding action according to the equipment status. It can be implemented using rule-based expert systems, fuzzy control algorithms, machine learning models (such as reinforcement learning, neural networks), etc. For example, it outputs the recommended feeding speed or mode based on a preset threshold or a trained model. Its main purpose is to achieve automated feeding control of the card crusher without human intervention.

[0076] Among them, the recommended feeding control action refers to the feeding operation recommended by the automated feeding control strategy generation logic in the current state. It can take the form of specific instructions such as feeding speed adjustment, feeding mode switching, and feeding pause. Its main purpose is to reflect the optimal control recommendations given by the automation system based on the current strategy.

[0077] Among them, the actual feeding control action refers to the feeding operation performed by the operator on the crusher through the control interface. It can take the form of specific instructions such as manual adjustment of feeding speed, emergency stop, switching to a specific mode, etc. It is mainly to reflect the operator's intervention in the control of the automation system based on experience.

[0078] Among them, corrective behavior refers to the situation where there is a difference between the actual feed control action performed by the operator and the recommended feed control action generated by the automation system. It is mainly to identify the operator's empirical correction to the automation control.

[0079] Among them, the observation time period refers to a period of time used to monitor the equipment operation data to evaluate the effect of the action after the operator performs the actual feed control action. It can be preset or dynamically determined based on the material characteristics and the type of corrective action. It is mainly for obtaining the data required to evaluate the effect of operator intervention.

[0080] Among them, multi-dimensional equipment performance improvement information refers to quantitative indicators for evaluating the impact of the operator's corrective behavior on the performance of the card crusher. It can be in the form of efficiency improvement indicators, energy consumption reduction indicators, stability enhancement indicators, etc. Its main purpose is to quantify the effectiveness of operator experience intervention.

[0081] Among them, the corrective feedback signal refers to the information that integrates the context when the operator's corrective behavior occurs, the behavior itself and the effects it brings. It is mainly used to provide a basis for adjusting the generation logic of the automated feeding control strategy.

[0082] The core innovation of this application is that by monitoring the actual feed control actions performed by the operator outside of the actions recommended by the automation system, the operator's corrective behavior is identified, and the equipment performance improvement information is monitored after the behavior is executed, the operator's empirical intervention and its effect are converted into a corrective feedback signal, and the feedback signal is used to dynamically adjust the automatic feed control strategy generation logic, which solves the problem that the existing system cannot learn from the operator's experience and has difficulty in dealing with complex materials, and achieves the effect of improving the performance and adaptability of the crusher.

[0083] Specifically, this method first acquires multi-dimensional equipment operating signals from the card shredder and extracts multi-dimensional state features from them to comprehensively perceive the equipment and material status. Then, based on these state features, the current automated feed control strategy generation logic generates a recommended feed control action. Simultaneously, the system monitors the operator's actual feed control actions. By comparing the operator's actual actions with the automated system's recommended actions, the operator's corrective actions are identified. Once a corrective action is identified, the system monitors the card shredder's equipment operating data for an observation period following the actual action and, based on this data, evaluates the multi-dimensional equipment performance improvement resulting from the corrective action. Next, the identified multi-dimensional state features, the recommended feed control actions, the operator's actual feed control actions, and the evaluated multi-dimensional equipment performance improvement information are integrated into a corrective feedback signal. Finally, based on this corrective feedback signal, the automated feed control strategy generation logic is adjusted. The adjusted strategy generation logic is then used for subsequent feed control of the card shredder. This entire process forms a closed loop, enabling the automated system to continuously learn from the operator's experience in complex operating conditions and incorporate this information into its own control strategy.

[0084] Through the above scheme, this application can identify and learn the operator's empirical feed control strategy in the process of complex and changeable material processing, and convert the operator's experience into the ability of the automation system, thereby overcoming the limitations of existing automation systems that are difficult to cope with complex working conditions, improving the processing efficiency, energy consumption performance and operating stability of the crusher when processing complex materials, and enhancing the adaptability of the equipment.

[0085] In some embodiments, step A1 comprises:

[0086] A101. Get the multi-dimensional device operation signal of the card crusher; the multi-dimensional device operation signal includes a motor current signal, a device sound signal, and a device vibration signal;

[0087] A102 performs waveform analysis on the motor current signal to extract impact load characteristics;

[0088] A103 performs spectrum analysis on the device sound signal to extract sound spectrum features;

[0089] A104. Perform spectrum analysis on the vibration signal of the device to extract vibration spectrum characteristics;

[0090] A105. Combine the impact load characteristics, the sound spectrum characteristics, and the vibration spectrum characteristics to form the multi-dimensional state characteristics.

[0091] Waveform analysis refers to the study and processing of the shape of a signal as it changes over time. This can be achieved through methods such as time-domain analysis, feature point detection, and instantaneous value calculation. Shock load characteristics are values ​​or patterns extracted from the motor current signal waveform that characterize the instantaneous impact or high stress experienced by a tool or device. These characteristics can be expressed through current spike amplitude, current rate of change, or current integral within a specific time window.

[0092] Among them, spectrum analysis refers to the technology of decomposing a signal into different frequency components, which can be achieved by methods such as fast Fourier transform (FFT), short-time Fourier transform (STFT), and wavelet analysis. Sound spectrum characteristics refer to the numerical values ​​or patterns extracted from the spectrum of the device sound signal, which can characterize the energy distribution of the device sound at different frequencies or specific frequency components. They can be expressed in terms of energy in a specific frequency band, resonance peak frequency, harmonic content, etc. Vibration spectrum characteristics refer to the numerical values ​​or patterns extracted from the spectrum of the device vibration signal, which can characterize the energy distribution of the device vibration at different frequencies or specific frequency components. They can be expressed in terms of vibration amplitude in a specific frequency band, main vibration frequency, harmonic vibration intensity, etc.

[0093] Among them, the multi-dimensional state feature refers to a set of numerical values ​​or vectors formed by combining features of different dimensions, which can comprehensively characterize the current operating state of the card shredder.

[0094] Based on the above technical features, the method of this application achieves its functions and solves the technical problem through the following means. First, by acquiring three dimensions of equipment operating signals: motor current, equipment sound, and equipment vibration. This method overcomes the limitation of relying solely on a single or simple signal and provides basic data for comprehensively perceiving the operating status of the card shredder. Next, waveform analysis of the motor current signal is performed to extract impact load characteristics, which can capture transient stress changes during material entry or crushing. Spectral analysis of the equipment sound signal is performed to extract acoustic spectrum characteristics, which can reflect auditory information about the interaction between the cutter and the material and the internal mechanical state of the equipment. Spectral analysis of the equipment vibration signal is performed to extract vibration spectrum characteristics, which can perceive the mechanical stability and balance of the equipment, as well as the vibration response caused by material impact. Through these refined analysis methods, the features extracted from different dimensions can more accurately and meticulously characterize the actual operating conditions of the card shredder when processing complex and variable materials, including material hardness, toughness, and shape, as well as potential issues such as wear and loosening of the equipment cutters. Finally, these impact load characteristics, acoustic spectrum characteristics, and vibration spectrum characteristics are combined to form a multi-dimensional state feature vector. This multi-dimensional state feature vector contains richer and more detailed state information than traditional methods. When this more refined multi-dimensional state feature is applied to the subsequent automated feed control strategy generation, operator corrective behavior identification, performance evaluation, and strategy adjustment processes, the automated system can more accurately understand the complex operating conditions of the current equipment, identify the corrective actions taken by the operator under specific subtle conditions and the resulting performance improvements, and thus more effectively learn the operator's experience strategy for handling complex materials. This learning and strategy adjustment mechanism based on multi-dimensional, refined state features enables the automated feed control strategy to better adapt to complex and changing material characteristics and equipment conditions, overcoming the control lag and poor control problems caused by insufficient state information in traditional systems, thereby improving the overall performance and adaptability of the crusher when handling complex materials.

[0095] In some embodiments, step A3 comprises:

[0096] A301. Get the actual control instructions and parameter settings executed by the operator through the card crusher control interface;

[0097] A302. Analyze the actual control instructions and parameter settings to identify the type and parameters of the actual feed control action;

[0098] A303. Get the type and parameters of the recommended feed control action;

[0099] A304. Compare the type and parameters of the actual feed control action with the type and parameters of the recommended feed control action to determine whether there is a difference between the actual feed control action and the recommended feed control action;

[0100] A305. When there is a difference, it is determined that there is a corrective action, and the actual feed control action is used as the operator's corrective action.

[0101] In the above steps, the actual control instructions and parameter settings executed by the operator through the shredder control interface are obtained. The shredder control interface can be any user interface or input device used by the operator to directly intervene in the shredder feeding process, such as a physical control panel, touch screen interface, remote monitoring terminal, or specific software application. The actual control instructions and parameter settings refer to the original operating commands and related values ​​or options entered by the operator through these interfaces, such as a "pause" command, a set feed speed value, or the option to select a specific feed mode.

[0102] The actual control instructions and parameter settings are parsed to identify the type and parameters of the actual feed control action. This parsing process converts the original, diverse instructions and parameters input by the operator into a standardized representation that the system can understand and process. The types of actual feed control actions may include, but are not limited to, start, stop, pause, resume, set speed, adjust torque limit, switch feed mode, etc. Parameters can be specific values ​​associated with these action types, such as set speed values, torque limit values, selected mode identifiers, etc. The type and parameters of the recommended feed control action are obtained. The recommended feed control action is generated by the automated feed control strategy generation logic based on the current state of the card shredder. The representation of its type and parameters corresponds to the representation of the actual feed control action to facilitate comparison.

[0103] The type and parameters of the actual feed control action are compared with those of the recommended feed control action to determine whether there are any differences between the two. This comparison process may include comparing whether the action types are consistent, and if the action types are consistent, comparing whether the relevant parameter values ​​are within the preset allowable deviation range. If there are differences, a corrective action is determined to have occurred, and the actual feed control action is used as the operator's corrective action. This determination is made based on the comparison results, and any inconsistency in type or parameters is considered an operator intervention in the automated recommendation, i.e., a corrective action. Using the actual feed control action itself as the corrective action means that the subsequent learning process will directly use the specific operations actually performed by the operator as learning samples.

[0104] Through the above steps, the method of the present application accurately identifies corrective operator actions. Specifically, by first capturing the operator's raw input on the control interface, the operator's detailed operational intent and specific values ​​are captured. These raw inputs are then parsed and standardized into action types and parameter representations understandable to the system. Simultaneously, the types and parameters of the recommended actions generated by the automated system are obtained. Next, the operator's actual actions are accurately compared with the recommended actions, identifying any discrepancies in action type or parameter values. Once a discrepancy is detected, it is determined to be a corrective action, and the actual action performed by the operator is used as the identification result. This method not only identifies simple speed adjustments but also captures more complex operator interventions such as pauses and mode switches. Compared to existing methods that only monitor changes in equipment status or simple speed adjustments, this method, by directly capturing and parsing the operator's raw control commands and parameters, more comprehensively and accurately reflects the operator's true operational intent and specific strategies. This precise identification of detailed operator interventions provides high-quality input data for the subsequent system to learn from the operator's experience, enabling more effective optimization of automated control strategies and improving the performance and adaptability of the card shredder when handling diverse and complex materials.

[0105] In some embodiments, step A4 comprises:

[0106] A401. When the operator's corrective action is identified, determining the observation period;

[0107] A402. During the observation period, monitor multi-dimensional equipment operating data for the card shredder; the multi-dimensional equipment operating data includes average motor power, motor current, equipment sound signal energy within a specific frequency range, equipment vibration signal amplitude within a specific frequency range, and the card shredder's discharge volume and discharge particle size distribution.

[0108] A403. Evaluate multi-dimensional equipment performance improvement information based on the multi-dimensional equipment operation data; the multi-dimensional equipment performance improvement information includes efficiency improvement indicators, energy consumption reduction indicators and stability enhancement indicators.

[0109] The observation time period may be determined based on factors such as the characteristics of the currently processed material, the type of corrective action taken by the operator, or changes in the state of the equipment, according to preset rules or a lookup table.

[0110] Among them, multi-dimensional equipment operation data refers to the signal or parameter set collected from the crusher that can reflect the operating status and effects of the equipment in multiple aspects. It can include electrical parameters reflecting motor load and stability, acoustic and vibration parameters reflecting the interaction between the tool and the material or potential abnormalities, and discharge parameters reflecting processing capacity and crushing quality.

[0111] Among them, the energy of the equipment sound signal within a specific frequency range refers to the total energy or average energy of the equipment operation sound signal within a preset frequency range related to specific physical phenomena (such as tool impact, material friction) (the specific frequency range can be predetermined by data statistics) after spectral analysis. It can be obtained by performing spectral analysis methods such as Fourier transform on the original sound signal, and integrating or averaging within the specified frequency range.

[0112] Among them, the amplitude of the equipment vibration signal within a specific frequency range refers to the vibration amplitude peak value or root mean square value within a preset frequency range related to a specific mechanical state (such as bearing wear and imbalance) (the specific frequency range can be predetermined by data statistics) after the equipment operation vibration signal undergoes spectral analysis. It can be obtained by performing spectral analysis on the original vibration signal and extracting characteristic values ​​within the specified frequency range.

[0113] Among them, the output particle size distribution refers to the distribution of material particle size after crushing by the crusher, which can be measured and characterized by screening analysis, image recognition or laser particle size analyzer.

[0114] The efficiency improvement index measures the degree of change in the crusher's processing capacity or effective output after operator intervention compared to before the intervention. For example, the discharge efficiency can be calculated by dividing the discharge volume during the observation period by the length of the observation period. The increase or increase ratio of the discharge efficiency after the intervention compared to the discharge efficiency before the intervention can then be calculated as the efficiency improvement index.

[0115] The energy consumption reduction index is an indicator that measures the degree of change in the energy consumption per unit of processing volume or per unit time of the card shredder after operator intervention, relative to the degree of change before the intervention. For example, the total energy consumption can be calculated by multiplying the average motor power by the length of the observation period. This total energy consumption is then divided by the output volume to obtain the energy consumption per unit of output volume. The energy consumption reduction index can then be calculated as the reduction in energy consumption per unit of output volume after the intervention, or the reduction ratio, relative to the energy consumption per unit of output volume before the intervention.

[0116] The stability enhancement index is an indicator that measures the degree of change in the operating stability of the crusher after operator intervention relative to before the intervention. The stability enhancement index can be a multi-dimensional indicator, which may include the standard deviation of the motor current (calculated based on the motor current), the coefficient of variation of the sound energy integral value (calculated based on the equipment sound signal energy within a specific frequency range), the coefficient of variation of the vibration amplitude peak value (calculated based on the equipment vibration signal amplitude within a specific frequency range), and the standard deviation of the discharge particle size distribution (calculated based on the discharge particle size distribution). The stability enhancement index can also be a weighted average of the normalized values ​​of these indicators.

[0117] After identifying the operator's corrective actions, this solution initiates the performance evaluation process. First, an appropriate observation period is determined. This sets clear time limits for subsequent data collection and evaluation, ensuring that the evaluation results are closely aligned with the operator's intervention. During this observation period, the system monitors multi-dimensional equipment operating data from the crusher. This data includes not only traditional average motor power, current, and discharge volume, but also more detailed and comprehensive parameters such as sound energy and vibration signal amplitude within specific frequency ranges, as well as discharge particle size distribution. This multi-dimensional data captures the equipment's true operating status and material handling performance under complex operating conditions, reflecting subtle changes that may result from operator intervention. Subsequently, based on this multi-dimensional equipment operating data, the system evaluates performance improvements across multiple dimensions, including efficiency improvement metrics, energy consumption reduction metrics, and stability enhancement metrics. This multi-dimensional evaluation approach quantifies the specific contributions of operator interventions in different areas. For example, an operator's adjustments may reduce energy consumption while maintaining high efficiency, or significantly improve stability at the expense of a small amount of efficiency. By correlating the operator's corrective actions, the equipment state at the time of the actions (reflected by the multi-dimensional state characteristics obtained in the previous steps), and the multi-dimensional performance improvement information after the actions, this solution can generate a high-quality corrective feedback signal. This feedback signal contains the essence of the operator's experience and can be used to adjust the logic for generating the automated feed control strategy, enabling the automated system to learn and reproduce the operator's effective intervention strategy under specific complex working conditions, thereby improving the adaptive ability and overall performance of the automated control. This mechanism of converting operator experience into learnable, multi-dimensional, quantifiable feedback information and using it to optimize the automation strategy is the key to solving the existing technical problems of this solution.

[0118] By monitoring multi-dimensional equipment operating data and evaluating multi-dimensional equipment performance improvement information based on this data, this solution can comprehensively and accurately capture the impact of the operator's corrective actions on the shredder's operating performance. Specifically, by monitoring multi-dimensional data such as the motor's average power, motor current, the energy of the equipment's sound signal within a specific frequency range, the amplitude of the equipment's vibration signal within a specific frequency range, the shredder's discharge volume, and the discharge particle size distribution, a detailed understanding of the equipment's actual operating status after operator intervention can be achieved. By evaluating multi-dimensional indicators such as efficiency improvement indicators, energy consumption reduction indicators, and stability enhancement indicators based on this data, the contribution of operator intervention in different aspects can be quantified. This multi-dimensional, refined evaluation overcomes the shortcomings of traditional methods that rely solely on a limited number of parameters. It can more accurately understand the value of operator experience strategies and provide high-quality feedback information for subsequent learning and adjustment of automation strategies, thereby improving the shredder's adaptability and overall performance when processing complex materials.

[0119] Preferably, step A401 may include:

[0120] Extracting state features reflecting characteristics of the currently processed material from the multi-dimensional state features as material characterization features;

[0121] The observation time period is determined according to the material characterization characteristics and the type of the actual feeding control action, according to a preset observation time period determination rule or an observation time period lookup table.

[0122] Among them, material characterization features refer to state features that are screened or extracted from multi-dimensional state features and are closely related to the physical properties of the material currently being processed (such as hardness, toughness, humidity, size, etc.). They can be determined in advance from multi-dimensional state features using principal component analysis, feature selection algorithms, or feature combinations based on expert knowledge.

[0123] Among them, the preset observation time period determination rule refers to the logic or algorithm established based on historical data analysis, machine learning model training or domain expert experience, which is used to calculate or derive the corresponding observation time period based on the input material characterization characteristics and the actual feed control action type. It can be implemented using a decision tree, regression model or a fuzzy logic-based reasoning system. The observation time period lookup table refers to a pre-established data structure that stores the recommended observation time periods corresponding to different material characterization characteristics (or their classifications) and different actual feed control action type combinations. It can be implemented using a two-dimensional or multi-dimensional array, hash table or database table. Determining the observation time period means calculating or obtaining a specific time length based on the current material characterization characteristics and the actual feed control action type by applying the preset determination rules or querying the preset lookup table, as the time window for subsequent monitoring of equipment operation data.

[0124] After identifying an operator's corrective action, this solution doesn't simply use a fixed observation period or one based solely on the type of action. Instead, it first identifies and extracts state features closely related to the material being processed from the multi-dimensional state features reflecting the equipment's real-time status. These features serve as material characterizations, thereby gaining an understanding of the material's state under the current operating conditions. Subsequently, based on the specific type of feed control action performed by the operator, the system determines rules or queries a pre-built observation period lookup table based on the pre-established observation period. These rules or lookup tables associate different material characteristics (reflected by the material characterizations) with different corrective action types and the corresponding effect onset time. For example, for fragile, dry materials, a slight operator deceleration may quickly result in reduced current fluctuations or a smoother sound. In this case, the rules or lookup table would define a shorter observation period. However, for tough, moist, or impure materials, or if the operator performs actions such as reversals or long pauses, stabilization or improvement of the equipment's state may take longer to manifest. In these cases, the rules or lookup table would define a longer observation period. In this way, the process of determining the observation time period fully considers the actual characteristics of the currently processed material and the nature of the operator's corrective actions, so that the determined time period can more accurately match the actual manifestation cycle of the impact of the corrective action on equipment performance. Limiting subsequent equipment operation data monitoring to such a more targeted observation time period can collect data that better reflects the actual effect of the corrective action, thereby laying a solid foundation for the subsequent accurate evaluation of multi-dimensional equipment performance improvement information. This dynamic and adaptive observation time period determination mechanism based on material characteristics and action type significantly improves the accuracy and effectiveness of operator experience learning, allowing the system to more accurately understand the reasons why operators take specific actions under specific complex working conditions and the actual effects they bring.

[0125] Preferably, step A403 may include:

[0126] B1. Calculate the evaluation values ​​of the efficiency improvement index, the energy consumption reduction index, and the stability enhancement index based on the multi-dimensional device operation data;

[0127] B2. Identify information reflecting the corrective behavior goal based on the actual feed control action, recorded as corrective intention information;

[0128] B3. Determine the weight coefficients of the efficiency improvement index, the energy consumption reduction index, and the stability enhancement index based on the material characterization characteristics and the correction intention information;

[0129] B4. Calculate the comprehensive performance improvement index based on the evaluation values ​​of the efficiency improvement index, the energy consumption reduction index, and the stability enhancement index and the corresponding weight coefficients;

[0130] B5. The evaluation values ​​of the efficiency improvement index, the energy consumption reduction index, the stability enhancement index, and the comprehensive performance improvement index are combined into the multi-dimensional equipment performance improvement information.

[0131] Among them, the calculation method of the evaluation values ​​of the efficiency improvement index, energy consumption reduction index and stability enhancement index can be referred to in the previous article.

[0132] Among them, the corrective intention information refers to the potential purpose or goal of the operator's intervention behavior inferred by analyzing the operator's actual feed control actions. It can be expressed by a preset intention category (for example, improving efficiency, reducing energy consumption, enhancing stability, avoiding jamming, optimizing discharge, etc.) or a more detailed intention description.

[0133] Among them, the comprehensive performance improvement index refers to an overall evaluation value calculated by combining the evaluation value of each performance indicator with its corresponding weight coefficient, which can be calculated using weighted summation, weighted average or other forms of aggregation functions.

[0134] This solution defines a detailed process for evaluating multi-dimensional equipment performance improvement information, aiming to more accurately and intelligently assess the effectiveness of operator corrective actions, thereby providing high-quality feedback for automated strategy learning. Specifically, based on multi-dimensional equipment operating data, evaluation values ​​for efficiency improvement, energy consumption reduction, and stability enhancement are calculated. This serves as the basis for quantitatively evaluating the effectiveness of operator corrective actions and directly reflects the equipment's performance in these key performance dimensions after the intervention. This calculation based on multi-dimensional equipment operating data ensures the objectivity and data support of the evaluation results. Furthermore, information reflecting the corrective action objectives is identified based on actual feed control actions and recorded as corrective intent information. This step is key to understanding operator experience. By analyzing the specific control action types and parameters performed by the operator, the system attempts to infer the operator's underlying motivation for the intervention, such as improving processing efficiency, reducing energy consumption, or enhancing equipment stability. Identifying corrective intent information enables more targeted performance evaluation and understands which performance objective or objectives the operator prioritizes in a specific context. Based on this, weighting coefficients for efficiency improvement, energy consumption reduction, and stability enhancement are determined based on material characterization and corrective intent information. This step further enhances the intelligent level of evaluation. It recognizes that the importance of various performance indicators varies when handling materials with different characteristics and when operators have different corrective actions. For example, when handling materials prone to sticking, stability may be more critical. If the operator's intention is to resolve an overload problem, energy consumption reduction and stability enhancement may be given higher weight. Specifically, based on the identified corrective action information and material characteristics, the weight coefficients for each performance indicator are determined by consulting a preset weight lookup table or applying a weight calculation rule. Dynamically adjusting the weights of each performance indicator based on the material characteristics (reflecting the characteristics of the material being processed) and the corrective action information ensures that the performance evaluation is more consistent with actual working conditions and the operator's experience, avoiding the shortcomings of simple averaging or fixed weighting. Subsequently, a comprehensive performance improvement index is calculated based on the evaluated values ​​of the efficiency improvement index, energy consumption reduction index, and stability enhancement index and their corresponding weight coefficients. This step combines the previously calculated evaluation values ​​of each performance indicator with the weights determined based on the material characteristics and operator intention to calculate a comprehensive evaluation index. This comprehensive index more comprehensively and accurately reflects the overall effectiveness of the operator's corrective actions, especially when balancing different performance objectives. Calculating a comprehensive performance improvement index provides a quantitative, context- and intent-based feedback signal for subsequent automated strategy learning. Finally, the evaluation values ​​of the efficiency improvement index, energy consumption reduction index, and stability enhancement index, along with the comprehensive performance improvement index, are combined to form multi-dimensional device performance improvement information. This step integrates the previously calculated evaluation values ​​and comprehensive indicators into a structured information package.This multi-dimensional equipment performance improvement information not only includes the evaluation of each raw performance, but also includes comprehensive evaluation results that take into account material and intention weights. It provides rich and valuable feedback information for the subsequent generation of corrective feedback signals and adjustment of automation strategies, enabling the automation system to more effectively learn from the operator's advanced experience.

[0135] By incorporating this more accurate and context-aware performance evaluation information into the adjustment process of the automated feed control strategy, the system can better understand the operator's decision-making logic under complex working conditions, thereby learning and replicating these advanced experiences, improving the adaptability and robustness of the automation strategy, and ultimately optimizing the overall performance of the crusher when processing diverse materials.

[0136] Preferably, step B2 may include:

[0137] Based on a preset intention recognition rule or an intention lookup table, the correction intention information is identified according to the type and parameters of the actual feed control action and the multi-dimensional state characteristics.

[0138] Among them, the preset intent recognition rules or intent lookup table refers to a pre-established knowledge structure used to map specific input combinations to corrective intent information. The rules can be a series of condition-action statements, and the lookup table can be a multi-dimensional mapping table. They are constructed based on the analysis of operator experience and equipment behavior.

[0139] This solution defines a specific implementation method for identifying corrective intent information, aiming to address the difficulty of accurately determining an operator's true intention based solely on their actual feed control actions. By combining the multi-dimensional state characteristics of the equipment at the time the operator performed the action, the operator's corrective action's purpose can be more accurately inferred. Upon identifying the operator's corrective action, the system obtains the multi-dimensional state characteristics at the time of the action. The type and parameters of the actual feed control action are combined with these multi-dimensional state characteristics as input. Based on pre-set intent recognition rules or an intent lookup table, this input is matched or inferred to identify the operator's true corrective intent for the action. For example, if an operator reduces the feed speed when a specific abnormal frequency appears in the equipment's vibration spectrum, the intent recognition rules may identify this as an intention to address a jam or protect a tool. This more accurate intent recognition enables subsequent performance evaluation weighting based on intent to more accurately reflect the operator's key performance indicators under specific operating conditions. For example, if the intent is identified as equipment protection, the stability enhancement indicator may be weighted higher when evaluating performance improvement; if the intent is identified as efficiency improvement, the efficiency improvement indicator may be weighted higher. In this way, this solution makes the performance improvement evaluation brought about by the operator's corrective behavior more consistent with the operator's real purpose and actual effect, thereby providing more accurate and instructive feedback signals for subsequent strategy adjustments, and ultimately improving the adaptability and performance of the automated feeding control strategy under complex working conditions.

[0140] In one embodiment, an intent lookup table can be constructed, whose inputs include the actual feeding action type, speed adjustment parameters, impact load characteristics, sound spectrum characteristics, and vibration spectrum characteristics. This lookup table can contain multiple entries. For example, one entry can map the combination of a "speed adjustment" action, a 20% speed reduction parameter, a high impact load characteristic, a normal sound spectrum characteristic, and a normal vibration spectrum characteristic to the corrective intent information for "protecting equipment." Another entry can map the combination of a "speed adjustment" action, a 10% speed reduction parameter, a moderate impact load characteristic, a specific high-frequency enhanced sound spectrum characteristic, and a normal vibration spectrum characteristic to the corrective intent information for "handling material hard spots." For another example, an entry can map the combination of a "pause" action, an inapplicable parameter, a moderate impact load characteristic, a normal sound spectrum characteristic, and a specific low-frequency enhanced vibration spectrum characteristic to the corrective intent information for "handling a stuck material." In actual application, when an operator performs an action, the system obtains the action type, parameters, and multi-dimensional state characteristics at that time, then queries the lookup table to find a matching entry and outputs the corresponding corrective intent information.

[0141] In some embodiments, step A6 includes:

[0142] A601. Based on the multi-dimensional state characteristics in the corrected feedback signal, locate the strategy parameters or rules corresponding to the multi-dimensional state characteristics in the automated feed control strategy generation logic as target strategy parameters or target rules;

[0143] A602. Calculate the feedback intensity of the actual feed control action based on the evaluation value of the efficiency improvement index, the evaluation value of the energy consumption reduction index, the evaluation value of the stability enhancement index, and the comprehensive performance improvement index in the multi-dimensional equipment performance improvement information in the corrected feedback signal;

[0144] A603. Adjust the target strategy parameters or target rules according to the feedback strength to increase the probability of generating the actual feed control action under a state similar to the multi-dimensional state characteristics.

[0145] The strategy parameters or rules of multi-dimensional state characteristics refer to the adjustable components of the automated feed control strategy generation logic. For example, in a rule-based system, these can be the rule conditions, priorities, or associated actions; in a machine learning model, these can be the model weights, biases, or other hyperparameters. The target strategy parameters or target rules refer to the strategy parameters or rules determined to require adjustment based on the current multi-dimensional state characteristics during a specific strategy adjustment process.

[0146] Among them, feedback intensity refers to a quantitative value calculated based on the degree of improvement in equipment performance brought about by the operator's actual feed control action. It is used to measure the effectiveness and importance of the operator's intervention. It can be calculated by a weighted combination of performance improvement indicators, a mapping based on a lookup table, or through a learning algorithm.

[0147] Among them, adjusting the target policy parameters or target rules to increase the probability of generating actual feed control actions under states similar to the multi-dimensional state characteristics refers to modifying the corresponding part of the automated feed control policy generation logic so that when encountering equipment states similar to those during the recorded operator intervention in the future, the automated system is more likely to generate recommended actions that are the same or similar to the corrective actions actually performed by the operator. This can be achieved by modifying the triggering conditions or priorities of the rules, adjusting the weights or biases of the model output layer, or updating the policy function in reinforcement learning.

[0148] By combining the aforementioned technical approaches, the method of the present application achieves the following operational principles: First, when the system recognizes that an operator has made a corrective intervention in an automated recommended action and generates a corrective feedback signal containing multi-dimensional state characteristics, actual actions, and performance improvement information, the method leverages the multi-dimensional state characteristics in the feedback signal to accurately identify the policy parameters or rules most relevant to that specific state within the complex automated feed control strategy generation logic. This ensures that the policy adjustments are tailored to the specific operating conditions in which the operator actually intervened, improving learning efficiency and accuracy. Next, the method calculates the feedback strength of the operator's actual feed control action based on the multi-dimensional equipment performance improvement information contained in the feedback signal, specifically quantifying the efficiency, energy consumption, stability, and comprehensive performance indicators that quantify the operator's intervention effect. The more significant the performance improvement, the greater the calculated feedback strength, indicating the more valuable the operator's experience. Finally, based on the calculated feedback strength, the method adjusts the target policy parameters or target rules identified previously. The magnitude of the adjustment is related to the feedback strength; the greater the feedback strength, the greater the adjustment. The goal of these adjustments is to increase the probability that the automation strategy will generate the same corrective action as the operator's actual execution when encountering future conditions similar to the recorded multi-dimensional state characteristics. This probabilistic adjustment mechanism, based on state similarity and the degree of performance improvement, enables the automation strategy to selectively and focusedly incorporate successful operator experiences, gradually improving its adaptability and control performance when handling complex materials and operating conditions.

[0149] This method makes full use of the rich and detailed feedback information provided by the previous steps (obtaining multi-dimensional state characteristics, identifying corrective behaviors, evaluating multi-dimensional performance improvements, and generating corrective feedback signals), so that strategy adjustments are no longer simple, blind attempts, but are based on a deep understanding of the specific context and behavioral effects of the operator's behavior. This allows for more effective learning from the operator's experience and wisdom, and improves the intelligence level of automated control.

[0150] To further illustrate the method of the present application, the following example is provided: Assume that the automated feed control strategy generation logic is a rule-based expert system containing a series of feed control rules triggered by equipment status. When the system recognizes that the operator has performed a corrective action under a specific multi-dimensional state characteristic, and that this action has resulted in a significant multi-dimensional improvement in equipment performance, a corrective feedback signal is generated. In step A601, based on the multi-dimensional state characteristic in the feedback signal, the expert system's rule base is searched for the rule or rule set that best matches the state characteristic, and these rules are identified as target rules. For example, if the state characteristic indicates a potential risk of material entanglement, and the operator briefly pauses the feed, the system may locate the feed rule associated with a "high risk of entanglement" state. In step A602, based on the evaluation values ​​of the efficiency improvement, energy consumption reduction, stability enhancement, and overall performance improvement indicators in the feedback signal, the feedback strength of the feed pause is calculated using a pre-set function or lookup table. For example, if the motor current fluctuation is significantly reduced, the sound spectrum returns to normal, and the overall performance indicators are significantly improved after the pause, the calculated feedback strength will be high. Step A603 can adjust the target rule located in step A601 based on the calculated high feedback intensity. This adjustment can be done by increasing the triggering priority of the target rule in similar situations, modifying the conditions of the rule to make it more likely to be triggered in similar situations, or associating a higher probability weight with the rule, thereby increasing the probability that the automated system will generate a recommended action to pause feeding when encountering similar entanglement risk situations in the future.

[0151] By adopting the method of the present application, the following technical effects can be achieved: by accurately locating the strategy adjustment target based on the multi-dimensional state characteristics in the corrective feedback signal, ineffective adjustments to irrelevant strategy parts are avoided, making strategy learning targeted. By calculating the feedback intensity based on multi-dimensional equipment performance improvement information, the value of operator intervention is quantified, ensuring that strategy learning is based on operator behavior that has a positive impact on equipment performance. By adjusting the target strategy parameters or rules based on the feedback intensity, the automation strategy can focus on and probabilistically learn the operator's successful experience, improving the adaptability and control performance when processing complex materials and coping with complex working conditions.

[0152] Preferably, step A602 may include:

[0153] Determining preliminary feedback intensity based on the comprehensive performance improvement indicator;

[0154] determining intention compliance based on the correction intention information, the evaluation value of the efficiency improvement index, the evaluation value of the energy consumption reduction index, and the evaluation value of the stability enhancement index;

[0155] The preliminary feedback strength is modified according to the intention compliance to obtain a final feedback strength.

[0156] The preliminary feedback intensity refers to the initial feedback amount calculated based on the overall performance improvement of the device, which can be achieved by converting the comprehensive performance improvement index into a basic intensity value through a mapping function (such as a linear function or a nonlinear function).

[0157] Among them, intention compliance refers to the degree of match between the various performance improvements brought about by the operator's actual operations and the operator's corrective intentions. It can be achieved by calculating the similarity or distance between the actual performance evaluation value vector and the reference performance improvement vector representing the operator's intentions (the reference performance improvement vector corresponding to each corrective intention information can be determined in advance).

[0158] The final feedback strength refers to the amount of feedback after adjustment for intent compliance, which is used to ultimately guide the adjustment of the automation strategy. This can be achieved by multiplying or adding the initial feedback strength to the intent compliance.

[0159] When calculating the feedback strength for actual feed control actions, this solution first calculates a preliminary feedback strength based on the observed improvement in overall equipment performance after the operator's corrective actions. This provides a basis for feedback based on actual results. Based on this, the operator's corrective intention is incorporated, combined with the observed performance evaluations of specific performance indicators (efficiency, energy consumption, and stability), to calculate an intention compliance score. This intention compliance quantifies the extent to which the operator's actual actions achieved their specific objectives. For example, if the operator's intention was to improve stability, and stability improved significantly after the actual operation, the intention compliance score is high. Finally, the calculated intention compliance score is used to modify the preliminary feedback strength. If the intention compliance score is high, it indicates that the operator's corrective actions were very effective in achieving their specific objectives. In this case, the feedback strength can be appropriately increased, making the automation strategy more likely to learn and adopt this behavior in similar situations. Conversely, if the intention compliance score is low, even if the overall performance indicators are acceptable, it may indicate that the behavior is not optimal for the operator. In this case, the feedback strength can be appropriately reduced to prevent the strategy from overlearning behaviors that are inconsistent with the operator's intentions. By combining the operator's corrective intent with actual performance improvements and using this information to adjust feedback intensity, this solution can more accurately quantify the value of the operator's corrective actions, enabling the automated feed control strategy generation logic to more effectively learn the operator's empirical preferences based on specific objectives in specific situations. This feedback mechanism, combined with operator intent, makes the strategy adjustment process more intelligent and targeted, better adapting to changing and complex material characteristics and working conditions, thereby improving the automation system's adaptability and optimization effectiveness.

[0160] refer to Figure 2 The present application provides a dynamic control system for feeding a plastic chip shredder, which is used to control the feeding action of the plastic chip shredder. The system includes:

[0161] State feature acquisition module 1, used to acquire the multi-dimensional device operation signal of the card shredder and extract the multi-dimensional state features therefrom (the specific process can be referred to step A1 above);

[0162] Recommended action generation module 2, for generating recommended feed control actions based on the multi-dimensional state characteristics using automated feed control strategy generation logic (for the specific process, please refer to step A2 above);

[0163] Corrective action recognition module 3 is used to monitor the actual feed control action performed by the operator on the card shredder, compare the actual feed control action with the recommended feed control action, and identify the operator's corrective action (the specific process can be referred to step A3 above);

[0164] Performance evaluation module 4 is used to monitor the operating data of the card crusher equipment during the observation period after the actual feed control action is executed when the operator's corrective action is identified, so as to evaluate the multi-dimensional equipment performance improvement information (the specific process can be referred to step A4 above);

[0165] Feedback signal generation module 5, used to generate a correction feedback signal based on the multi-dimensional state characteristics, the recommended feed control action, the actual feed control action and the multi-dimensional equipment performance improvement information (the specific process can be referred to step A5 above);

[0166] Strategy adjustment module 6, used to adjust the automatic feeding control strategy generation logic according to the correction feedback signal (the specific process can be referred to step A6 above);

[0167] The control execution module 7 is used to generate logic according to the adjusted automatic feeding control strategy to control the feeding of the card shredder (the specific process can be referred to the above step A7).

[0168] The above description is merely an embodiment of the present application and is not intended to limit the scope of protection of the present application. For those skilled in the art, various modifications and variations of the present application are possible. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A dynamic control method for feeding a plastic chip shredder, used to control the feeding action of the plastic chip shredder, characterized in that: The steps of the method include: A1. Obtain multi-dimensional device operation signals from the card shredder and extract multi-dimensional status features from them; A2. Generate recommended feed control actions based on the multi-dimensional state characteristics using automated feed control strategy generation logic; A3. Monitoring the actual feed control action performed by the operator to crush the card machine, comparing the actual feed control action with the recommended feed control action to identify the operator's corrective behavior; A4. When an operator's corrective action is identified, monitor the card shredder equipment operating data during the observation period after the actual feed control action is executed to evaluate multi-dimensional equipment performance improvement information; A5. Generate a corrective feedback signal based on the multi-dimensional state characteristics, the recommended feed control action, the actual feed control action, and the multi-dimensional device performance improvement information; A6. Adjust the automated feed control strategy generation logic according to the corrective feedback signal; A7. Generate logic based on the adjusted automated feeding control strategy to control the feeding of the card crusher.

2. A dynamic control method for feeding a card crusher according to claim 1, characterized in that: Step A1 includes: A101. Get the multi-dimensional device operation signal of the card crusher; the multi-dimensional device operation signal includes a motor current signal, a device sound signal, and a device vibration signal; A102 performs waveform analysis on the motor current signal to extract impact load characteristics; A103 performs spectrum analysis on the device sound signal to extract sound spectrum features; A104. Perform spectrum analysis on the vibration signal of the device to extract vibration spectrum characteristics; A105. Combine the impact load characteristics, the sound spectrum characteristics, and the vibration spectrum characteristics to form the multi-dimensional state characteristics.

3. A dynamic control method for feeding a card crusher according to claim 1, characterized in that: Step A3 includes: A301. Get the actual control instructions and parameter settings executed by the operator through the card crusher control interface; A302. Analyze the actual control instructions and parameter settings to identify the type and parameters of the actual feed control action; A303. Get the type and parameters of the recommended feed control action; A304. Compare the type and parameters of the actual feed control action with the type and parameters of the recommended feed control action to determine whether there is a difference between the actual feed control action and the recommended feed control action; A305. When there is a difference, it is determined that there is a corrective action, and the actual feed control action is used as the operator's corrective action.

4. A card crusher feed dynamic control method according to claim 3, characterized in that: Step A4 includes: A401. When the operator's corrective action is identified, determining the observation period; A402. During the observation period, monitor multi-dimensional equipment operating data for the card shredder; the multi-dimensional equipment operating data includes average motor power, motor current, equipment sound signal energy within a specific frequency range, equipment vibration signal amplitude within a specific frequency range, and the card shredder's discharge volume and discharge particle size distribution. A403. Evaluate multi-dimensional equipment performance improvement information based on the multi-dimensional equipment operation data; the multi-dimensional equipment performance improvement information includes efficiency improvement indicators, energy consumption reduction indicators and stability enhancement indicators.

5. A dynamic control method for feeding a card crusher according to claim 4, characterized in that: Step A401 includes: Extracting state features reflecting characteristics of the currently processed material from the multi-dimensional state features as material characterization features; The observation time period is determined according to the material characterization characteristics and the type of the actual feeding control action, according to a preset observation time period determination rule or an observation time period lookup table.

6. A dynamic control method for feeding a card crusher according to claim 5, characterized in that: Step A403 includes: B1. Calculate the evaluation values ​​of the efficiency improvement index, the energy consumption reduction index, and the stability enhancement index based on the multi-dimensional device operation data; B2. Identify information reflecting the corrective behavior goal based on the actual feed control action, recorded as corrective intention information; B3. Determine the weight coefficients of the efficiency improvement index, the energy consumption reduction index, and the stability enhancement index based on the material characterization characteristics and the correction intention information; B4. Calculate the comprehensive performance improvement index based on the evaluation values ​​of the efficiency improvement index, the energy consumption reduction index, and the stability enhancement index and the corresponding weight coefficients; B5. The evaluation values ​​of the efficiency improvement index, the energy consumption reduction index, the stability enhancement index, and the comprehensive performance improvement index are combined into the multi-dimensional equipment performance improvement information.

7. A dynamic control method for feeding a card crusher according to claim 6, characterized in that: Step B2 includes: Based on a preset intention recognition rule or an intention lookup table, the correction intention information is identified according to the type and parameters of the actual feed control action and the multi-dimensional state characteristics.

8. A dynamic control method for feeding a card crusher according to claim 6, characterized in that: Step A6 includes: A601. Based on the multi-dimensional state characteristics in the corrected feedback signal, locate the strategy parameters or rules corresponding to the multi-dimensional state characteristics in the automated feed control strategy generation logic as target strategy parameters or target rules; A602. Calculate the feedback intensity of the actual feed control action based on the evaluation value of the efficiency improvement index, the evaluation value of the energy consumption reduction index, the evaluation value of the stability enhancement index, and the comprehensive performance improvement index in the multi-dimensional equipment performance improvement information in the corrected feedback signal; A603. Adjust the target strategy parameters or target rules according to the feedback strength to increase the probability of generating the actual feed control action under a state similar to the multi-dimensional state characteristics.

9. A dynamic control method for feeding a card crusher according to claim 8, characterized in that: Step A602 includes: Determining preliminary feedback intensity based on the comprehensive performance improvement indicator; determining intention compliance based on the correction intention information, the evaluation value of the efficiency improvement index, the evaluation value of the energy consumption reduction index, and the evaluation value of the stability enhancement index; The preliminary feedback strength is modified according to the intention compliance to obtain a final feedback strength.

10. A dynamic control system for feeding plastic chips into a shredder, used to control the feeding action of the plastic chip shredder, characterized in that: The system includes: A state feature acquisition module is used to obtain the multi-dimensional device operation signal of the card shredder and extract multi-dimensional state features from it; A recommended action generation module, configured to generate a recommended feed control action based on the multi-dimensional state characteristics using an automated feed control strategy generation logic; a corrective action identification module, configured to monitor actual feed control actions performed by the operator on the card shredder, and compare the actual feed control actions with the recommended feed control actions to identify the operator's corrective actions; a performance evaluation module for monitoring the card crusher equipment operating data during an observation period after the actual feed control action is executed when the operator's corrective action is identified, so as to evaluate multi-dimensional equipment performance improvement information; a feedback signal generating module, configured to generate a corrective feedback signal based on the multi-dimensional state characteristics, the recommended feed control action, the actual feed control action, and the multi-dimensional equipment performance improvement information; A strategy adjustment module, configured to adjust the automatic feeding control strategy generation logic according to the correction feedback signal; The control execution module is used to generate logic according to the adjusted automatic feeding control strategy to control the feeding of the card crusher.

Images