A self-adaptive intelligent control system and control method for a ceramic dry granulator

Abstract

This invention discloses an adaptive intelligent control system and control method for a ceramic dry granulation machine, belonging to the field of industrial automation and intelligent manufacturing technology. The method includes the following steps: S1, offline modeling and rule initialization; S2, real-time sensing and material state identification; S3, dynamic working condition judgment and strategy triggering; S4, adaptive adjustment strategy execution; S5, quality closed-loop verification and strategy self-optimization; S6, full-process linkage and safety protection. This invention is rationally designed, achieving high-precision and stable control of particle size distribution, morphology, and bulk density through multi-source sensor fusion, online material characteristic identification, and multi-variable collaborative adjustment. It is suitable for dry granulation scenarios involving no liquid phase participation, high friction, and easy material blockage.

Description

Technical Field

[0001] This invention relates to the field of industrial automation and intelligent manufacturing technology, and more specifically to the field of an adaptive intelligent control system and control method for a dry ceramic granulator. Background Technology

[0002] Dry granulation of ceramics offers significant advantages in energy saving and consumption reduction because it eliminates the need for a drying process. However, the process is highly dependent on the equipment's real-time response to material conditions: changes in raw material flowability (such as humidity fluctuations or an increase in the proportion of fine powder) can easily lead to material blockage, overload, or uneven particle size. Currently, most domestic and international granulators operate with fixed parameters or use simple PID feedback control, which cannot dynamically adapt to changes in material characteristics.

[0003] Patent CN110420673A, entitled "A Microfluidic Device and Driving Method Thereof, Microfluidic System," discloses the following: A microfluidic device and its driving method, and a microfluidic system, used to solve the problem of droplets having difficulty crossing electrode gaps. The embodiment of this application provides a microfluidic device comprising: a first substrate and a second substrate placed opposite each other; a droplet-containing space is formed between the first substrate and the second substrate; the first substrate includes: an array of first electrodes, and the second substrate includes: an array of second electrodes; in at least one arrangement direction of the first electrodes, the gaps between the first electrodes and the gaps between the second electrodes do not overlap; the first electrodes and the second electrodes are configured to: apply a voltage to the first electrodes and the second electrodes to control the movement of droplets between the first substrate and the second substrate.

[0004] The patent with publication number CN120754762A and patent name "Ceramic Dry Granulation Unit" discloses the following: It includes a granulator body and a first horizontal plate fixed to its middle outer wall. The bottom and top of the granulator body are respectively equipped with a cleaning mechanism and a conveying mechanism. This invention relates to the field of granulation equipment technology. The ceramic powder is transported to the right by a screw feeder, and at the same time, a second fan blows air into the discharge cylinder, so that the ceramic powder can be sprayed out from the distributor more quickly and evenly. At this time, the external water supply device injects water into several water pipes, and the water is sprayed out through atomizing spray heads to wet the ceramic powder sprayed from the distributor. Part of the wet ceramic powder will fall directly onto the turntable through the bottom opening of the granulator body, and part will fall onto the inner wall of the bottom of the granulator body. At this time, the first fan blows air into the inner cavity of the bottom of the granulator body, thereby blowing up the wet ceramic powder that has fallen onto the inner wall of the bottom of the granulator body, so that it falls onto the turntable by its own weight.

[0005] Patent CN110420673A only implements sequential start-stop control; patent CN120754762A focuses on mechanical structure improvement and does not involve intelligent regulation. While imported equipment (such as Gerteis) has a certain degree of self-adaptive capability, its control logic is closed and its cost is high.

[0006] Therefore, there is an urgent need for an open, customizable, and domestically produced intelligent control system with material sensing and strategy optimization capabilities to solve the problems of unstable particle quality, poor equipment reliability, and low level of intelligent control in existing technologies. Summary of the Invention

[0007] The purpose of this invention is to solve the technical problems of unstable particle quality, poor equipment reliability, and low degree of intelligent control in the prior art. This invention provides an adaptive intelligent control system and control method for a ceramic dry granulation machine.

[0008] To achieve the above objectives, the present invention specifically adopts the following technical solution: This invention provides an adaptive intelligent control method for a ceramic dry granulation machine, comprising the following steps: S1. Offline modeling and rule initialization: Based on historical batch data, a "particle quality - process parameter" mapping model is established using multiple regression or neural network methods. The basic setpoint ranges of spindle speed N, extrusion pressure P, and feed rate F corresponding to different target particle sizes are generated and stored in the PLC experience library as the initial control benchmark. S2. Real-time sensing and material status identification: The extrusion pressure, material temperature, spindle speed, feed rate, and main motor load current of the granulator are collected in real time through multi-source sensing units; each sensing signal is isolated and transmitted to the PLC controller. The PLC controller performs sliding window filtering and outlier removal on the raw data, and extracts "feed flow rate standard deviation σ_F" and "main motor load current change rate ΔI / Δt" as material flowability characteristics. S3. Dynamic operating condition judgment and strategy triggering: The PLC controller executes the judgment logic once at a certain time interval: if σ_F is greater than the set value and ΔI / Δt>the set value, it is determined that the current material has entered the "poor flowability" operating condition, the adaptive adjustment flag is immediately triggered, and the corresponding strategy combination in the experience library is matched by looking up the table; otherwise, the conventional PID control is maintained. S4. Adaptive Adjustment Strategy Execution: Based on the judgment result of step S3, the PLC controller dynamically adjusts key parameters through the actuator unit. Under normal operating conditions: maintain PID closed-loop control to keep the spindle speed, extrusion pressure, and feed rate within the target range; In cases of poor liquidity: The PLC controller calls the corresponding strategy combination from the experience base and synchronously issues control commands: Reduce the spindle speed by 5% to 10% and adjust the main motor frequency using a frequency converter; Increase the extrusion pressure to the set upper limit, which is adjusted via the proportional valve of the hydraulic system; Reduce the feed screw frequency to the lower limit by adjusting the frequency converter of the feed metering feeder; S5. Quality Closed-Loop Verification and Strategy Self-Optimization: The PLC controller indirectly infers the particle density through the steady-state current I_ss and torque T_est of the main motor. If I_ss deviates from the benchmark value by ±10% for several consecutive cycles, the "quality abnormality" flag is triggered, and the system automatically rolls back to the strategy for the poor flowability condition in step S4. If the quality returns to normal after adjustment, the strategy execution path, material characteristics, and adjustment effect are archived to the experience library for rapid matching and strategy enhancement for similar conditions in the future. S6. Full-process linkage and safety protection: The PLC controller automatically detects the status of the dust removal fan and cooling water valve before the host starts; during operation, it monitors the motor load in real time. If the current suddenly increases to 150% of the rated value and continues for more than the set time, it is determined that the material is blocked, and the machine is immediately stopped and reversed to clean it.

[0009] In one implementation, in step S1, based on the historical batch target particle size D data, the following particle quality-process parameter mapping model is constructed: The target particle size D50 ∈ [0.8mm, 1.2mm] corresponds to the following parameter ranges for spindle speed N, extrusion pressure P, and feed rate F: Spindle speed N∈[120,180rpm]; Extrusion pressure P∈[8,12MPa]; The feed rate F ∈ [1.5, 2.5 t / h].

[0010] In one implementation, in step S3, the PLC controller collects process data every 500ms and executes the following judgment logic: Calculate the standard deviation of the feed flow rate σ_F; Calculate the rate of change of the main motor load current ΔI / Δt; If the following conditions are met: σ_F>5% and ΔI / Δt>8A / s, then it is determined as "poor material flowability".

[0011] In one implementation, step S5 involves the following closed-loop verification of particle quality: S51. The particle density is indirectly deduced by the steady-state current I_ss and torque T_est of the main motor. S52. If I_ss deviates from the benchmark value by ±10% for three consecutive periods, the "quality anomaly" flag is triggered, and the strategy of poor liquidity condition in step S4 is automatically rolled back. S53. Effective adjustment strategies are automatically archived for quick recall under similar operating conditions.

[0012] In one implementation, in step S6, when a sudden increase in motor current to 150% of the rated value is detected and lasts for more than 2 seconds, the granulator is triggered to stop due to material blockage.

[0013] In one implementation, in step S42, the control strategy is stored in the PLC experience base and supports automatic matching according to raw material batches. The construction of the PLC experience base includes the following process: A1. Historical Data Acquisition and Preprocessing: Collect granulator operation data under different material characteristics (such as moisture content, particle size distribution, and flowability grade) in multiple batches, including parameters such as extrusion pressure, spindle speed, feed rate, main motor load current, torque, and particle density, and perform preprocessing such as normalization, noise reduction, and outlier removal on the data. A2. Feature Extraction and Working Condition Classification: Feature extraction is performed on the preprocessed data to extract key feature variables, such as "standard deviation of feed flow rate", "rate of change of motor load current" and "trend of material temperature change", as the basis for judging the "poor flowability" working condition; and typical working condition intervals are divided according to different material types (such as high moisture material, fine powder material, recycled material, etc.). A3. Regulation strategy encoding and rule generation: By combining expert experience with on-site commissioning data, a mapping rule for "operating condition - adjustment strategy" is formed; for example: If "poor fluidity" is triggered, the following strategy combination will be invoked: [reduce rotation speed by 5%~10%, increase pressure to the upper limit, and reduce feed frequency to the lower limit]; If "quality anomaly" is triggered consecutively, the policy rollback mechanism will be activated. A4. Offline training and optimization of the experience base: Offline simulation verification of strategy combinations is carried out using historical data and simulation platforms to optimize adjustment range and trigger threshold, avoiding over-adjustment or oscillation; similar working conditions are merged through cluster analysis or decision tree methods to improve the generalization ability of the experience base. A5. Online update mechanism for the experience base: The PLC controller records the input characteristics and adjustment effects of each strategy execution during actual operation. If the adjustment effect is good under the same working condition multiple times in a row, the weight of the strategy is strengthened; if the adjustment effect is poor (such as the particle quality not being restored), the strategy replacement or parameter fine-tuning is triggered, and the adjustment log is retained for subsequent manual review and model iteration.

[0014] A second aspect of the present invention provides an adaptive intelligent control system for a ceramic dry granulator, applicable to the aforementioned adaptive intelligent control method for a ceramic dry granulator, comprising a multi-source sensing unit, an actuator unit, a central control unit, and a human-machine interface, wherein the multi-source sensing unit, the actuator unit, and the human-machine interface are signal-connected to the central control unit.

[0015] In one embodiment, the multi-source sensing unit is used to collect material status and equipment operating parameters in real time. The multi-source sensing unit includes: A pressure sensor, installed in the extrusion chamber of the granulator, is used for pressure detection in closed-loop regulation of the extrusion pressure; The PT100 temperature sensor is installed at the discharge port of the granulator for monitoring the material temperature. An encoder, installed on the main shaft of the granulator, is used to feed back the rotational speed to the central control unit; Weighing belt, used for real-time measurement of feed flow rate; Current transformers are installed in the main motor circuit for current monitoring and quality assessment. A flow meter, used in conjunction with a water weighing system, is used for monitoring the influent flow rate.

[0016] Specifically, the pressure sensor is installed on the side wall of the granulation chamber, and its feedback value is used for PID adjustment of the extrusion pressure in step S4. The encoder is installed at the end of the spindle, and its pulse signal is used to calculate the real-time speed of the spindle and serves as the speed closed-loop feedback in step S4. The current transformer collects the three-phase current of the main motor, which is used to calculate ΔI / Δt in step S2 and to evaluate the particle density in step S5.

[0017] In one embodiment, the actuator unit is used to receive instructions from the central control unit and perform physical adjustments, including: The frequency converter is used to drive the main motor and realize the adjustment of the spindle speed; A quantitative feeder is used to adjust the feeding rate; The water weighing system adjusts the actuator, which works in conjunction with the flow meter to control the water-to-material ratio. Hydraulic proportional valves are used for precise adjustment of extrusion pressure.

[0018] In one embodiment, the central control unit includes: The PLC controller has a built-in adaptive control module and experience strategy library for data processing, working condition identification, strategy matching, instruction issuance, quality verification and safety protection. Human-machine interface: a touch screen, supporting target particle size setting, real-time curve display, alarm recording, and manual correction based on experience database.

[0019] Specifically, the PLC controller's internal experience strategy library supports automatically loading corresponding control strategies according to material batch numbers and records the effect of each strategy execution for model iteration; The touchscreen has a built-in historical data playback function, which allows operators to trace the changes in operating conditions and control response processes for each batch.

[0020] The beneficial effects of this invention are as follows: 1. The adaptive intelligent control method for ceramic dry granulation machine of the present invention is based on a dedicated adaptive intelligent control system. The system completes data acquisition through multi-source sensing units, completes core control through central control unit, completes action execution through actuator unit, and completes human-machine interaction through human-machine interface. The units work together to achieve adaptive intelligent control of the granulation process.

[0021] 2. This invention achieves high-precision and stable control of particle size distribution, morphology and bulk density by multi-source sensor fusion, online identification of material characteristics and multi-variable collaborative adjustment. It is suitable for dry granulation scenarios with no liquid phase participation, high friction and easy material blockage.

[0022] 3. An adaptive intelligent control system and method with online material identification, multi-variable decoupling adjustment, and closed-loop verification of quality and effect capabilities, to achieve "supply control based on quality and adjustment based on demand", thereby improving product consistency and equipment robustness. Attached Figure Description

[0023] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained from these drawings without creative effort.

[0024] Figure 1 This is a schematic diagram of an adaptive intelligent control system for a ceramic dry granulator according to the present invention. Detailed Implementation

[0025] To make the technical problems, technical solutions, and technical effects of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0026] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0027] Example 1 This invention provides an adaptive intelligent control method for a ceramic dry granulation machine, comprising the following steps: S1. Offline modeling and rule initialization: Based on historical batch data, a "particle quality - process parameter" mapping model is established using multiple regression or neural network methods. The basic setpoint ranges of spindle speed N, extrusion pressure P, and feed rate F corresponding to different target particle sizes are generated and stored in the PLC experience library as the initial control benchmark. Construct the following particle quality-process parameter mapping model: The target particle size D50 ∈ [0.8mm, 1.2mm] corresponds to the following parameter ranges for spindle speed N, extrusion pressure P, and feed rate F: Spindle speed N∈[120,180rpm]; Extrusion pressure P∈[8,12MPa]; The feed rate F ∈ [1.5, 2.5 t / h].

[0028] The construction of the PLC experience base includes the following process: A1. Historical Data Acquisition and Preprocessing: Collect granulator operation data under different material characteristics (such as moisture content, particle size distribution, and flowability grade) in multiple batches, including parameters such as extrusion pressure, spindle speed, feed rate, main motor load current, torque, and particle density, and perform preprocessing such as normalization, noise reduction, and outlier removal on the data. A2. Feature Extraction and Working Condition Classification: Feature extraction is performed on the preprocessed data to extract key feature variables, such as "standard deviation of feed flow rate", "rate of change of motor load current" and "trend of material temperature change", as the basis for judging the "poor flowability" working condition; and typical working condition intervals are divided according to different material types (such as high moisture material, fine powder material, recycled material, etc.). A3. Regulation strategy encoding and rule generation: By combining expert experience with on-site commissioning data, a mapping rule for "operating condition - adjustment strategy" is formed; for example: If "poor fluidity" is triggered, the following strategy combination will be invoked: [reduce rotation speed by 5%~10%, increase pressure to the upper limit, and reduce feed frequency to the lower limit]; If "quality anomaly" is triggered consecutively, the policy rollback mechanism will be activated. A4. Offline training and optimization of the experience base: Offline simulation verification of strategy combinations is carried out using historical data and simulation platforms to optimize adjustment range and trigger threshold, avoiding over-adjustment or oscillation; similar working conditions are merged through cluster analysis or decision tree methods to improve the generalization ability of the experience base. A5. Online update mechanism for the experience base: The PLC controller records the input characteristics and adjustment effects of each strategy execution during actual operation. If the adjustment effect is good under the same working condition multiple times in a row, the weight of the strategy is strengthened; if the adjustment effect is poor (such as the particle quality not being restored), the strategy replacement or parameter fine-tuning is triggered, and the adjustment log is retained for subsequent manual review and model iteration.

[0029] S2. Real-time sensing and material status identification: The extrusion pressure, material temperature, spindle speed, feed rate, and main motor load current of the granulator are collected in real time through multi-source sensing units; each sensing signal is isolated and transmitted to the PLC controller. The PLC controller performs sliding window filtering and outlier removal on the raw data, and extracts "feed flow rate standard deviation σ_F" and "main motor load current change rate ΔI / Δt" as material flowability characteristics. S3. Dynamic operating condition judgment and strategy triggering: The PLC controller executes the judgment logic every 500ms: if σ_F>5% and ΔI / Δt>8A / s, it is determined that the current material has entered the "poor flowability" operating condition, the adaptive adjustment flag is immediately triggered, and the corresponding strategy combination in the experience library is matched by looking up the table; otherwise, the conventional PID control is maintained. S4. Adaptive Adjustment Strategy Execution: Based on the judgment result of step S3, the PLC controller dynamically adjusts key parameters through the actuator unit. Under normal operating conditions: maintain PID closed-loop control to keep the spindle speed, extrusion pressure, and feed rate within the target range; In cases of poor liquidity: The PLC controller calls the corresponding strategy combination from the experience base and synchronously issues control commands: Reduce the spindle speed by 5% to 10% and adjust the main motor frequency using a frequency converter; Increase the extrusion pressure to the set upper limit, which is adjusted via the proportional valve of the hydraulic system; Reduce the feed screw frequency to the lower limit by adjusting the frequency converter of the feed metering feeder; S5. Quality Closed-Loop Verification and Strategy Self-Optimization: The PLC controller indirectly infers the particle density through the steady-state current I_ss and torque T_est of the main motor. If I_ss deviates from the benchmark value by ±10% for three consecutive cycles, the "quality abnormality" flag is triggered, and the strategy for the poor flowability condition in step S4 is automatically rolled back. If the quality returns to normal after adjustment, the execution path, material characteristics, and adjustment effect of this strategy are archived to the experience library for rapid matching and strategy enhancement for similar conditions in the future. S6. Full-process linkage and safety protection: The PLC controller automatically detects the status of the dust removal fan and cooling water valve before the host starts; during operation, it monitors the motor load in real time. If the current suddenly increases to 150% of the rated value and lasts for more than 2 seconds, it is determined that the material is blocked, and the machine is immediately stopped and reversed to clean it.

[0030] Table 1 shows the comparative test results on a certain ceramic pilot line. As can be seen from Table 1, the particle uniformity is significantly improved, the equipment stability is enhanced, and the overall energy saving is about 11.6%.

[0031] Table 1 Comparative Test Results on a Certain Ceramic Pilot Line

[0032] Example 2 This embodiment provides an adaptive intelligent control system for a ceramic dry granulation machine, including a multi-source sensing unit, an actuator unit, a central control unit, and a human-machine interface. The multi-source sensing unit, actuator unit, and human-machine interface are signal-connected to the central control unit.

[0033] The multi-source sensing unit is used to collect material status and equipment operating parameters in real time, including a pressure sensor, a PT100 temperature sensor, an encoder, a weighing belt, a current transformer, and a flow meter. The pressure sensor is installed in the extrusion chamber of the granulator for pressure detection in closed-loop regulation of the extrusion pressure. The PT100 temperature sensor is installed at the discharge port of the granulator for material temperature monitoring. The encoder is installed on the main shaft of the granulator for feedback of rotational speed to the central control unit. The weighing belt is used for real-time measurement of feed flow. The current transformer is installed in the main motor circuit for current monitoring and quality assessment. The flow meter works with the water weighing system for monitoring the inlet water flow.

[0034] Specifically, a pressure sensor is installed on the side wall of the granulation chamber, and its feedback value is used for PID adjustment of the extrusion pressure in step S4; an encoder is installed at the end of the main shaft, and its pulse signal is used to calculate the real-time speed of the main shaft and as a closed-loop speed feedback in step S4; a current transformer collects the three-phase current of the main motor, which is used to calculate ΔI / Δt in step S2 and to evaluate the particle density in step S5.

[0035] The actuator unit is used to receive instructions from the central control unit and perform physical adjustments. It includes a frequency converter, a quantitative feeder, a water weighing system regulating actuator, and a hydraulic proportional valve. The frequency converter is used to drive the main motor and realize the adjustment of the spindle speed. The quantitative feeder is used to adjust the feeding rate. The water weighing system regulating actuator works with a flow meter to realize the water-to-material ratio control. The hydraulic proportional valve is used for precise adjustment of the extrusion pressure.

[0036] The central control unit includes a PLC controller and a touch screen for human-machine interaction. The PLC controller has a built-in adaptive control module and experience strategy library for data processing, working condition judgment, strategy matching, instruction issuance, quality verification and safety protection. The touch screen for human-machine interaction supports target particle size setting, real-time curve display, alarm recording and manual correction based on experience library.

[0037] Specifically, the PLC controller's internal experience strategy library supports automatic loading of corresponding control strategies based on material batch numbers and records the effect of each strategy execution for model iteration; it also includes a monitoring and control system connected to the PLC controller signals, which includes an operator station, an engineer station, and a redundant server.

[0038] The touchscreen has a built-in historical data playback function, which allows operators to trace the changes in operating conditions and control response processes for each batch.

[0039] The core of the adaptive intelligent control in this embodiment lies in the closed-loop logic of "perception (multi-source sensing unit) - judgment - adjustment (actuator) - verification - optimization": real-time perception of material and equipment status is achieved through multi-source sensing unit, intelligent judgment of working conditions is achieved through PLC controller algorithm, multi-parameter collaborative adaptive adjustment is achieved through actuator unit, effectiveness verification of control strategy is achieved through particle quality closed-loop verification, and control strategy is archived and optimized through PLC experience library, ultimately realizing unmanned, intelligent adaptive control of granulation process.

Claims

1. An adaptive intelligent control method for a ceramic dry granulation machine, characterized in that, Includes the following steps: S1. Offline modeling and rule initialization: Based on historical batch data, a "particle quality - process parameter" mapping model is established using multiple regression or neural network methods. The basic setpoint ranges of spindle speed N, extrusion pressure P, and feed rate F corresponding to different target particle sizes are generated and stored in the PLC experience library as the initial control benchmark. S2. Real-time sensing and material status identification: The extrusion pressure, material temperature, spindle speed, feed rate, and main motor load current of the granulator are collected in real time through multi-source sensing units; each sensing signal is isolated and transmitted to the PLC controller. The PLC controller performs sliding window filtering and outlier removal on the raw data, and extracts "feed flow rate standard deviation σ_F" and "main motor load current change rate ΔI / Δt" as material flowability characteristics. S3. Dynamic operating condition judgment and strategy triggering: The PLC controller executes the judgment logic once at a certain time interval: if σ_F is greater than the set value and ΔI / Δt>the set value, it is determined that the current material has entered the "poor flowability" operating condition, the adaptive adjustment flag is immediately triggered, and the corresponding strategy combination in the experience library is matched by looking up the table; otherwise, the conventional PID control is maintained. S4. Adaptive Adjustment Strategy Execution: Based on the judgment result of step S3, the PLC controller dynamically adjusts key parameters through the actuator unit. Under normal operating conditions: maintain PID closed-loop control to keep the spindle speed, extrusion pressure, and feed rate within the target range; In cases of poor liquidity: The PLC controller calls the corresponding strategy combination from the experience base and synchronously issues control commands: Reduce the spindle speed by 5% to 10% and adjust the main motor frequency using a frequency converter; Increase the extrusion pressure to the set upper limit, which is adjusted via the proportional valve of the hydraulic system; Reduce the feed screw frequency to the lower limit by adjusting the frequency converter of the feed metering feeder; S5. Quality Closed-Loop Verification and Strategy Self-Optimization: The PLC controller indirectly infers the particle density through the steady-state current I_ss and torque T_est of the main motor. If I_ss deviates from the benchmark value by ±10% for several consecutive cycles, the "quality abnormality" flag is triggered, and the strategy for the poor flowability condition in step S4 is automatically rolled back. If the quality returns to normal after adjustment, the execution path, material characteristics, and adjustment effect of this strategy are archived to the experience library for rapid matching and strategy enhancement for similar conditions in the future. S6. Full-process linkage and safety protection: The PLC controller automatically detects the status of the dust removal fan and cooling water valve before the host starts; during operation, it monitors the motor load in real time. If the current suddenly increases to 150% of the rated value and continues for more than the set time, it is determined that the material is blocked, and the machine is immediately stopped and reversed to clean it.

2. The adaptive intelligent control method for a ceramic dry granulator according to claim 1, characterized in that, In step S1, based on the historical batch target particle size D data, the following particle quality-process parameter mapping model is constructed: The target particle size D50 ∈ [0.8mm, 1.2mm] corresponds to the following parameter ranges for spindle speed N, extrusion pressure P, and feed rate F: Spindle speed N∈[120,180rpm]; Extrusion pressure P∈[8,12MPa]; The feed rate F ∈ [1.5, 2.5 t / h].

3. The adaptive intelligent control method for a ceramic dry granulation machine according to claim 1, characterized in that, In step S3, the PLC controller collects process data every 500ms and executes the following judgment logic: Calculate the standard deviation of the feed flow rate σ_F; Calculate the rate of change of the main motor load current ΔI / Δt; If the following conditions are met: σ_F>5% and ΔI / Δt>8A / s, then it is determined as "poor material flowability".

4. The adaptive intelligent control method for a ceramic dry granulator according to claim 1, characterized in that, In step S5, the particle quality closed-loop verification is as follows: S51. The particle density is indirectly deduced by the steady-state current I_ss and torque T_est of the main motor. S52. If I_ss deviates from the benchmark value by ±10% for three consecutive periods, the "quality anomaly" flag is triggered, and the strategy of poor liquidity condition in step S4 is automatically rolled back. S53. Effective adjustment strategies are automatically archived for quick recall under similar operating conditions.

5. The adaptive intelligent control method for a ceramic dry granulator according to claim 1, characterized in that, In step S6, when the motor current suddenly increases to 150% of the rated value and lasts for more than 2 seconds, the granulator is triggered to stop due to material blockage.

6. The adaptive intelligent control method for a ceramic dry granulator according to claim 1, characterized in that, In step S42, the control strategy is stored in the PLC experience base, which supports automatic matching by raw material batch and manual correction and version management. The construction of the PLC experience base includes the following process: A1. Historical Data Acquisition and Preprocessing: Collect granulator operation data under different material characteristics in multiple batches, including extrusion pressure, spindle speed, feed rate, main motor load current, torque, and particle density parameters, and perform normalization, noise reduction, and outlier removal preprocessing on the data. A2. Feature Extraction and Working Condition Classification: Feature extraction is performed on the preprocessed data to extract key feature variables as the basis for judging the "poor flowability" working condition; and typical working condition intervals are divided according to different material types. A3. Regulation strategy encoding and rule generation: By combining expert experience with on-site commissioning data, a mapping rule for "operating condition - adjustment strategy" is formed; A4. Offline training and optimization of the experience base: Offline simulation verification of strategy combinations is carried out using historical data and simulation platforms to optimize adjustment range and trigger threshold, avoiding over-adjustment or oscillation; similar working conditions are merged through cluster analysis or decision tree methods to improve the generalization ability of the experience base. A5. Online update mechanism for the experience base: During actual operation, the PLC controller records the input characteristics and adjustment effects of each strategy execution. If the adjustment effect is good under the same working condition multiple times in a row, the weight of the strategy is strengthened; if the adjustment effect is poor, the strategy replacement or parameter fine-tuning is triggered, and the adjustment log is retained for subsequent manual review and model iteration.

7. An adaptive intelligent control system for a ceramic dry granulator, applicable to the adaptive intelligent control method for a ceramic dry granulator according to any one of claims 1 to 6, characterized in that, It includes a multi-source sensing unit, an actuator unit, a central control unit, and a human-machine interface. The multi-source sensing unit, actuator unit, and human-machine interface are connected to the central control unit via signals.

8. The adaptive intelligent control system for a ceramic dry granulation machine according to claim 7, characterized in that, The multi-source sensing unit is used to collect material status and equipment operating parameters in real time. The multi-source sensing unit includes: A pressure sensor, installed in the extrusion chamber of the granulator, is used for pressure detection in closed-loop regulation of the extrusion pressure; The PT100 temperature sensor is installed at the discharge port of the granulator for monitoring the material temperature. An encoder, installed on the main shaft of the granulator, is used to feed back the rotational speed to the central control unit; Weighing belt, used for real-time measurement of feed flow rate; Current transformers are installed in the main motor circuit for current monitoring and quality assessment. A flow meter, used in conjunction with a water weighing system, is used for monitoring the influent flow rate.

9. The adaptive intelligent control system for a ceramic dry granulation machine according to claim 7, characterized in that, The actuator unit is used to receive commands from the central control unit and perform physical adjustments, including: The frequency converter is used to drive the main motor and realize the adjustment of the spindle speed; A quantitative feeder is used to adjust the feeding rate; The water weighing system adjusts the actuator, which works in conjunction with the flow meter to control the water-to-material ratio. Hydraulic proportional valves are used for precise adjustment of extrusion pressure.

10. The adaptive intelligent control system for a ceramic dry granulator according to claim 7, characterized in that, The central control unit includes: The PLC controller has a built-in adaptive control module and experience strategy library for data processing, working condition identification, strategy matching, instruction issuance, quality verification and safety protection. Human-machine interface: a touch screen, supporting target particle size setting, real-time curve display, alarm recording, and manual correction based on experience database.

Images