Intelligent stirring control method and system

Abstract

The invention discloses an intelligent stirring control method and system, and relates to the field of stirring systems. The intelligent stirring control method comprises the following steps: S1, system parameter configuration and self-inspection; s2, material introduction and basic state monitoring; s3, performing dynamic viscosity evaluation and initial parameter matching; s4, stirring parameter dynamic feedback adjustment; s5, quantitatively adding the medicament; and S6, discharging the materials and checking the state. According to the intelligent stirring control method and system, initial parameters are preset in five viscosity intervals, viscosity evaluation is corrected in real time in combination with an LSTM model, the problem that traditional fixed parameters cannot adapt to wide-viscosity materials of 0-2000 mPas or above is solved, low-viscosity solutions, high-viscosity paste and the like can be accurately controlled, and the application scene is expanded.

Description

Technical Field

[0001] This invention relates to the field of stirring system technology, specifically to an intelligent stirring control method and system. Background Technology

[0002] In mixing processes in industries such as chemical, food, and pharmaceutical, the viscosity of materials varies greatly, and the parameters of the mixing process are highly coupled. Traditional control methods rely on experience to set fixed parameters, which presents the following challenges: First, high-viscosity materials are prone to causing excessive torque and equipment overload, while low-viscosity materials are prone to affecting the uniformity of mixing due to uneven flow field. Second, the parameter adjustment lacks dynamic feedback and cannot adapt to real-time changes in material properties; Third, there is a coupling contradiction between torque and uniformity optimization, and adjusting a single parameter is prone to neglecting one aspect. Although existing technologies attempt to introduce feedback control, they mostly focus on single-dimensional parameters (such as adjusting only the speed) and have not formed a synergistic correction mechanism for torque, uniformity, and reagent addition, making it difficult to ensure a balance between stirring effect and equipment safety under complex working conditions. Summary of the Invention

[0003] To address the shortcomings of existing technologies, this invention provides an intelligent stirring control method and system that solves the problems of poor parameter adaptability, the contradiction between torque and uniformity optimization, and insufficient response to abnormal working conditions in the stirring of complex materials.

[0004] To achieve the above objectives, the present invention provides the following technical solution: an intelligent stirring control method, comprising the following steps: S1. System parameter configuration and self-test: preset basic thresholds and adjustment parameters, perform functional verification of the sensor, and load the pre-trained LSTM viscosity model; S2. Material introduction and basic status monitoring: Perform the material introduction operation and simultaneously collect basic status parameters such as initial liquid level and temperature. S3. Viscosity dynamic evaluation and initial parameter matching: Collect dynamic and static parameters, evaluate the material viscosity in real time through the LSTM model and match it to the corresponding range, and call the initial stirring parameters of that range. S4. Dynamic feedback adjustment of stirring parameters: Parameters are adjusted based on real-time feedback of torque and uniformity, and cross-correction logic is executed. S5. Quantitative addition of reagents: Perform quantitative addition of reagents based on viscosity assessment results and stirring status; S6. Material Discharge and Status Verification: Execute the discharge operation, monitor the status parameters during the discharge process, and trigger a new cycle after the discharge is completed.

[0005] Preferably, in the S1 system parameter configuration and self-test step: The basic thresholds include the upper and lower limits of liquid level (10% to 90%), the torque safety threshold, the uniformity standard (≥85%), and the rotational speed and blade angle corresponding to each viscosity range; The formula for calculating the torque safety threshold is as follows: ,in: This refers to the motor's rated maximum torque, expressed in N·m. This represents the current material density, in kg / m³. The density is the standard material density, in kg / m³, e.g., 1000 for clean water; The adjustment parameters include the single adjustment range of the rotational speed (5% to 15%), the single adjustment range of the blade angle (2-10°), and the feedback verification interval (30 seconds to 1 minute). Sensor calibration must ensure that the data acquisition frequency of sensors such as torque, turbidity, and temperature is ≥10Hz, and the measurement error is ≤±2% (full scale).

[0006] Preferably, the rotational speed and blade angle corresponding to each viscosity range are: Range 1 (0-100 mPa・s): Rotation speed 300-500 r / min, blade angle 15-20°; Range 2: 101-500 mPa·s: rotational speed 501-800 r / min, blade angle 21-30°; Range 3: 501-1000 mPa·s; rotational speed 801-1200 r / min; blade angle 31-45°. Range 4: 1001-2000 mPa・s: rotational speed 1201-1500 r / min, blade angle 46-60°; Range 5, above 2001 mPa·s: rotational speed 1501-1800 r / min, blade angle 61-75°.

[0007] Preferably, in the S2 material introduction and basic status monitoring step, the basic status parameters collected include: Initial liquid level height, in meters, with an accuracy of ±0.01 meters; Initial temperature of the material, in °C, with an accuracy of ±0.5 °C; Import rate, measured in m³ / min, is recorded in real time via a flow sensor.

[0008] Preferably, in the S3 viscosity dynamic evaluation and initial parameter matching step: The dynamic parameters include real-time torque T, motor current, liquid level fluctuation value and stirring speed n, and the unit of real-time torque T is N·m, the unit of motor current is A, the unit of liquid level fluctuation value is m, and the unit of stirring speed n is r / min. Static parameters include the real-time temperature of the material. and material density , The unit is ℃; The viscosity evaluation steps are as follows: Preliminary estimate: Where K is the equipment characteristic coefficient, and it is taken as 0.5 for propulsion equipment and 0.8 for turbine equipment. For rotational speed, and The unit is d is the blade diameter, and the unit is m; LSTM Model Correction: Input Parameters such as initial temperature and density are used to output the final viscosity value. And the confidence level. If the confidence level is ≥90%, it is adopted directly; if the confidence level is <90%, it is adjusted based on historical data.

[0009] Preferably, in the S4 dynamic feedback adjustment step of stirring parameters: The torque feedback adjustment logic includes: Level 1 response: Basic speed reduction adjustment, when real-time torque... At that time, a level 1 response is triggered: According to the formula

[0010] in: This is the deceleration adjustment coefficient, ranging from 0.1 to 0.15. The higher the viscosity, the larger the value, making it suitable for high-resistance scenarios. The current stirring speed, in units of ; The new stirring speed after adjustment, in units of ; After the speed reduction is implemented, a 30-second delay is performed for verification; other parameter adjustments are paused to focus on monitoring the torque change trend and avoid oscillations caused by frequent adjustments. If after Level 1 response Enter stable maintenance mode: Maintain current speed and angle parameters, trigger fluctuation monitoring every 2 minutes (collect torque standard deviation). threshold Determined to be stable); like A limited number of further speed reductions will be implemented, with a cumulative speed reduction of [amount missing]. The original value should be used to avoid over-adjustment that could worsen uniformity. Secondary response: Angle coordination adjustment, if after the primary response This triggers a level 2 response; Adjust the blade angle by 5°-10° and simultaneously start a 30-second verification cycle to monitor the torque drop. If the torque is still Perform extreme speed reduction and angle readjustment: reduce the speed to the lowest speed in the viscosity range (e.g., viscosity range 3 corresponds to the lower limit of 800 r / min, which needs to be combined with the specific range parameters); further reduce the angle by 5° and start a 60-second countdown protection to forcibly reserve a safety buffer time; When the countdown ends Triggering limit protection: The system alarms and forces the speed down to the minimum safe speed (fixed value 200r / min, to ensure the transition state before the motor stops). If the torque drops back to a safe range After two consecutive successful verifications, the parameter recovery process was initiated. Speed ​​recovery: Verify the torque by increasing the speed by 5% each time (e.g., 200 r / min → 210 r / min → …) with a 30-second interval. Angle recovery: It is performed in increments of 2° each time (e.g., 40°→42°→…), in coordination with speed recovery, to avoid sudden changes in a single parameter.

[0011] Preferably, the uniformity feedback adjustment logic includes: Uniformity calculation formula:

[0012] in: Turbidity deviation refers to the turbidity deviation at different points within the mixing tank, which is collected in real time by a multi-channel sensor. and These represent the maximum and minimum turbidity values ​​monitored during the stirring cycle; when When the time is right, a level one response is triggered; if after two consecutive adjustments... This triggers a level 2 response; First-level response: Differentiated adjustment across viscosity ranges: Low viscosity range adjustment strategies for ranges 1 and 2: Prioritize increasing the rotational speed to enhance fluid disturbance, and perform "increase rotational speed by 10% + 1 minute verification".

[0013] If verified Maintain the current speed for 3 minutes, then drop it back to the upper limit of the range (e.g., the upper limit of 800 r / min corresponds to range 2) to balance energy consumption and uniformity. If verified And the foam is abnormal (the turbidity sensor identifies the foam signal): switch to "angle compensation + speed reduction", that is, increase the angle by 3°-5° and reduce the speed by 5% (suppress the foam while maintaining shear force); High viscosity range adjustment strategy (range 3-5): Enhance material shearing through angle optimization; dynamically calculate the new angle using a formula:

[0014] in: This is the angle adjustment coefficient, ranging from 5° to 10°. The higher the viscosity, the larger the value, making it suitable for high-resistance materials. This is the current blade angle; The blade angle has been adjusted. Execute "adjust to larger angle" "+1 minute verification": like Maintain the current angle parameters to ensure mixing stability; like : Superimpose “angle + speed coordinated adjustment”, that is, further increase the angle by 3°-5° + increase the speed by 5% (strengthen the coupling between shear and fluid motion); Level 2 Response: Coordinated Compensation of Agents and Time After two consecutive Level 1 adjustments This triggers a level 2 response: Addition of pesticide: Addition of pesticide in step S5 of the quantitative pesticide addition process; Time compensation: Extend the stirring time by 10% to 20%, and implement it simultaneously with the addition of the reagent to ensure thorough mixing; After adjustment, re-verify U; if it rises back to If the parameters fall back to normal, or if they still do not meet the requirements, a cross-correction logic is triggered.

[0015] Preferably, the cross-correction logic specifically includes the following hierarchical correction strategy: I. Low viscosity range correction scenario for ranges 1 and 2: Scenario 1: Reducing speed to maintain torque leads to a decrease in uniformity; When speed reduction is implemented due to excessive torque, and the uniformity U decreases by ≥10%, such as from 90% to ≤80%, a correction is triggered. Speed ​​compensation: Increase the speed to 90% to 95% of the speed before deceleration (strictly not exceeding the upper limit of the original range, such as 800 r / min for the original upper limit of range 2, then ≤760 r / min after compensation). Angle coordination: Simultaneously increase the blade angle by 2°-3° (to enhance shear force compensation and improve mixing efficiency); Validate closed loop: After compensation, maintain uniformity monitoring for 1 minute. If Then keep the current parameters; If still This triggers an extension of the mixing time (extending the base mixing time by 10% to compensate for homogenization through time). Scenario 2: Increasing speed to improve uniformity causes torque to approach the threshold (stop speed + angle / time adjustment); due cause After increasing the engine speed, the torque rises to (Not exceeding the limit but close to the threshold), triggering correction: Pause speed adjustment: Immediately stop any further speed increase operations; Angle priority: Increase the blade angle by 3°-5° (to improve uniformity by increasing the material contact area, thus mitigating the torque risk associated with high speed). Time compensation: Simultaneously extend the stirring time by 5% (in conjunction with angle adjustment to ensure mixing effect); Verification cycle: Monitor torque and uniformity every 2 minutes; if torque drops back to... and Restore normal adjustment logic; II. High viscosity range correction scenario in interval 3-5; Scenario 1: Increasing the angle to improve uniformity causes a sudden increase in torque (angle adjustment + slight decrease in engine speed). due cause After adjusting the angle, the torque surged to [a certain value] within 30 seconds. Triggering a fix: Emergency Reversal: The angle is immediately reversed to 90% of its original value (e.g., if the original angle is 45°, it is reversed to 40.5°), while the rotational speed decreases by 3% to 5% (quickly reducing the equipment load). Step-by-step verification: After the pullback, monitor the torque for 1 minute. If it stabilizes... Adjust the angle in steps of 50% of the original adjustment range (each step is 20 seconds apart; for example, if the original plan was to adjust by 5°, adjust by 2.5° each time to avoid torque fluctuations). Reassessment Trigger: If torque remains after callback Forced viscosity reassessment (re-execute the LSTM model and verify viscosity assessment bias).

[0016] Scenario 2: Slowing down to maintain torque results in consistently low uniformity (angle priority + reagent assistance) After two consecutive decelerations, the torque stabilizes at... but Triggering a fix: Angle enhancement: Based on the current rotational speed, according to the formula Increase the angle ( This is the original angle adjustment coefficient, such as the original... =10° then new =5°, reducing the risk of overload in the adjustment range); Step-by-step verification: Monitor uniformity 30 seconds after each angle increase. If U rises back to... Then maintain; if the angle has reached the upper limit of the interval (e.g., the upper limit of interval 5 is 75°), trigger: Pharmaceutical adjuvant: Add 3% to 5% of the drug; Time compensation: Extend stirring time by 15% (to synergistically improve homogenization effect with the agent).

[0017] Universal extreme scenario correction scenario across the entire range; Scenario 1: Dual failures continue to worsen, torque and ; When neither torque nor uniformity improves after two consecutive adjustments, an extreme correction is triggered. Safe speed reduction: Force speed reduction to 80% of the lowest speed in the range (e.g., if the lowest speed in range 3 is 800 r / min, then reduce to 640 r / min), while adjusting the blade angle to the median of the range (e.g., if the median angle in range 3 is 38°, to balance the load and mixing). Mixed mode: Switches between "low speed high angle (maintained for 60 seconds) → medium speed low angle (maintained for 60 seconds)" every 2 minutes (e.g., interval 3: 640r / min + 38° → 1000r / min + 33°) to break up material agglomeration through alternating flow fields; Manual warning: If the standard is not met after 5 minutes, an audible and visual alarm and a remote notification will be triggered to prompt manual inspection of the material properties, sensor status, or equipment mechanical failure.

[0018] Scenario 2: Parameter adjustment cumulative limit exceeded (speed / angle adjustment range too large) If the cumulative speed adjustment is ≥30% or the cumulative angle adjustment is ≥20° within 10 minutes, a correction will be triggered. Historical comparison: The system automatically retrieves the best parameter combination in the same viscosity range (efficient operating parameters stored in the database). Deviation verification: If the current parameter deviates from the historical best by ≥20%, execute parameter reset: Speed ​​recovery: The speed gradually decreases to the historical benchmark speed in increments of 5% each time (torque is verified at 1-minute intervals). Angle correction: Each time, the angle gradually decreases to the historical reference angle in increments of 2° (in coordination with speed correction to avoid sudden changes in a single parameter). Restart Adjustment: After resetting, restart the normal torque and uniformity parameter adjustment logic to ensure that the parameters are dynamically adapted within the optimal range.

[0019] An intelligent stirring control system is used to execute an intelligent stirring control method, comprising: The system parameter configuration and self-test module, with its preset basic thresholds and adjustment parameters, performs functional verification of the sensor and loads a pre-trained LSTM viscosity model. The material import and basic status monitoring module performs the material import operation and simultaneously collects basic status parameters such as initial liquid level and temperature. The viscosity dynamic assessment and initial parameter matching module collects dynamic and static parameters, evaluates the material viscosity in real time through the LSTM model and matches it to the corresponding range, and calls the initial stirring parameters for that range. The stirring parameter dynamic feedback adjustment module adjusts parameters based on real-time feedback of torque and uniformity, and executes cross-correction logic. The reagent dosing module performs the quantitative addition of reagents based on viscosity assessment results and stirring status. The material discharge and status verification module performs the discharge operation, monitors the status parameters of the discharge process, and triggers a new cycle after the discharge is completed.

[0020] Its beneficial effects are as follows: 1. This intelligent stirring control method and system solves the problem that traditional fixed parameters cannot be adapted to materials with a wide viscosity range of 0-2000 mPa·s by presetting initial parameters in 5 viscosity ranges and combining them with LSTM model to correct viscosity evaluation in real time. It can achieve precise control for low viscosity solutions, high viscosity pastes and other materials, and expand the applicable scenarios.

[0021] 2. The intelligent stirring control method and system, with its torque feedback-based first-level speed reduction, second-level angle adjustment, and limit protection stratified response mechanism, can prevent torque over-limit during the stirring of high-viscosity materials; the uniformity feedback-based interval adjustment strategy (adjusting speed for low viscosity and adjusting angle for high viscosity) can steadily improve the stirring uniformity, while reducing parameter oscillations through a 30-second / 1-minute verification cycle to ensure system stability.

[0022] 3. The intelligent stirring control method and system, with its cross-correction logic, addresses multiple typical coupled scenarios such as "reducing speed to ensure safety but sacrificing uniformity" and "adjusting angle to improve uniformity but causing a sudden increase in torque." Through a coordinated strategy of speed compensation, angle fine-tuning, and reagent assistance, it can maintain stable operation even under extreme conditions (such as double non-compliance or parameter cumulative over-limit), thus improving the adaptability of stirring complex materials. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a schematic diagram of the overall process of the present invention; Figure 2 This is a schematic diagram of the dynamic feedback adjustment steps for stirring parameters according to the present invention; Figure 3 This is a schematic diagram of the torque feedback judgment process of the present invention; Figure 4 This is a schematic diagram of the uniformity feedback judgment process of the present invention; Figure 5 This is a flowchart illustrating the cross-correction logic of the present invention. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention are described clearly and completely. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific implementation methods.

[0027] This invention discloses an intelligent stirring control method and system, according to the appendix. Figure 1 As shown, it includes the following steps: S1. System parameter configuration and self-test: preset basic thresholds and adjustment parameters, perform functional verification of the sensor, and load the pre-trained LSTM viscosity model; S2. Material introduction and basic status monitoring: Perform the material introduction operation and simultaneously collect basic status parameters such as initial liquid level and temperature. S3. Viscosity dynamic evaluation and initial parameter matching: Collect dynamic and static parameters, evaluate the material viscosity in real time through the LSTM model and match it to the corresponding range, and call the initial stirring parameters of that range. S4. Dynamic feedback adjustment of stirring parameters: Parameters are adjusted based on real-time feedback of torque and uniformity, and cross-correction logic is executed. S5. Quantitative addition of reagents: Perform quantitative addition of reagents based on viscosity assessment results and stirring status; S6. Material Discharge and Status Verification: Execute the discharge operation, monitor the status parameters during the discharge process, and trigger a new cycle after the discharge is completed.

[0028] In the quantitative dosing of medicines, the formula for calculating the dosage is as follows: ,in (The higher the viscosity, the larger the value), S is the bottom area of ​​the container (m²), and h is the liquid level height (m). This refers to the required concentration of the reagent (L / m³).

[0029] Material discharge and status verification in progress: Monitoring discharge rate , liquid level change For time; When the discharge rate is 30% lower than the historical average, mark "potential residue" and trigger the cleaning procedure; After the cleaning process is complete, perform sensor zero-point calibration.

[0030] Through a closed-loop process of parameter configuration, status monitoring, viscosity assessment, dynamic adjustment, reagent addition, and discharge circulation, intelligent control of the entire process of mixing materials with different viscosities is achieved, solving the problems of reliance on experience and poor parameter adaptability in traditional methods, and improving mixing efficiency and stability.

[0031] According to the appendix Figure 1 As shown, the S1 system parameter configuration and self-test steps are as follows: The basic thresholds include the upper and lower limits of liquid level (10% to 90%), the torque safety threshold, the uniformity standard (≥85%), and the rotational speed and blade angle corresponding to each viscosity range; The formula for calculating the torque safety threshold is as follows: ,in: This refers to the motor's rated maximum torque, expressed in N·m. This represents the current material density, in kg / m³. The density is the standard material density, in kg / m³, e.g., 1000 for clean water; The adjustment parameters include the single adjustment range of the rotational speed (5% to 15%), the single adjustment range of the blade angle (2-10°), and the feedback verification interval (30 seconds to 1 minute). Sensor calibration must ensure that the data acquisition frequency of sensors such as torque, turbidity, and temperature is ≥10Hz, and the measurement error is ≤±2% (full scale).

[0032] By using quantified basic thresholds, dynamically adapted torque safety thresholds, and high-precision sensor calibration, an accurate benchmark is provided for subsequent stirring control, avoiding adjustment failures caused by parameter ambiguity or sensor errors, and ensuring safe operation of the equipment (torque not exceeding limits) and stirring accuracy (uniformity meeting standards).

[0033] The corresponding rotational speed and blade angle for each viscosity range are as follows: Range 1 (0-100 mPa・s): Rotation speed 300-500 r / min, blade angle 15-20°; Range 2: 101-500 mPa·s: rotational speed 501-800 r / min, blade angle 21-30°; Range 3: 501-1000 mPa·s; rotational speed 801-1200 r / min; blade angle 31-45°. Range 4: 1001-2000 mPa・s: rotational speed 1201-1500 r / min, blade angle 46-60°; Range 5, above 2001 mPa·s: rotational speed 1501-1800 r / min, blade angle 61-75°.

[0034] By presetting parameters for different viscosity ranges, differentiated initial matching of materials with different viscosities can be achieved (low viscosity emphasizes rotation speed, high viscosity emphasizes angle), solving the problem that traditional fixed parameters cannot adapt to a wide viscosity range.

[0035] In the S2 material introduction and basic condition monitoring step, the basic condition parameters collected include: Initial liquid level height, in meters, with an accuracy of ±0.01 meters; Initial temperature of the material, in °C, with an accuracy of ±0.5 °C; Import rate, measured in m³ / min, is recorded in real time via a flow sensor.

[0036] By collecting high-precision basic parameters, data support is provided for viscosity assessment (temperature affects viscosity), reagent dosing (liquid level determines volume), and anomaly prediction (introduction rate reflects the initial state of the material), reducing the blindness of subsequent adjustments and improving control accuracy.

[0037] In the S3 viscosity dynamic evaluation and initial parameter matching steps: The dynamic parameters include real-time torque T, motor current, liquid level fluctuation value and stirring speed n, and the unit of real-time torque T is N·m, the unit of motor current is A, the unit of liquid level fluctuation value is m, and the unit of stirring speed n is r / min. Static parameters include the real-time temperature of the material. and material density , The unit is ℃; The viscosity evaluation steps are as follows: Preliminary estimate: Where K is the equipment characteristic coefficient, and it is taken as 0.5 for propulsion equipment and 0.8 for turbine equipment. For rotational speed, and The unit is d is the blade diameter, and the unit is m; LSTM Model Correction: Input Parameters such as initial temperature and density are used to output the final viscosity value. And the confidence level. If the confidence level is ≥90%, it is adopted directly; if the confidence level is <90%, it is adjusted based on historical data.

[0038] By employing a two-stage evaluation approach—preliminary estimation using physical formulas and intelligent correction using LSTM models—combining dynamic and static parameters, the problem of large viscosity errors in single-parameter evaluation is solved, thereby improving the accuracy of viscosity assessment and providing a reliable basis for parameter matching.

[0039] The specific process for correcting the LSTM model is as follows: The following multi-dimensional parameters are collected as the input vector X of the LSTM model: Preliminary viscosity estimate: using physical formulas Calculated; initial temperature of material Accuracy ±0.5℃, collected in real time by a temperature sensor; Real-time density of materials Detected online using a near-infrared component analyzer, with an accuracy of ±0.01 g / cm³; Stirring system status parameters: current stirring speed n (r / min), motor current I (A), standard deviation of liquid level fluctuation. (Unit: m, reflecting material flowability).

[0040] The input vector is organized into a time series. The time step is a sliding window of data from the first 30 seconds of the mixing process (covering changes in the initial state of the material). The LSTM model structure and training logic are as follows: A viscosity prediction model is constructed using a 3-layer LSTM network (including an input layer, hidden layers, and an output layer): Input layer: The dimension matches the input vector X (6-dimensional parameters), and the feature representation is enhanced by the ReLU activation function; Hidden layer: Set 64 LSTM units to learn the time dependence between parameters (such as the hysteresis effect of temperature change on viscosity). Output layer: Outputs two results, one is the corrected viscosity value. (Unit: mPa·s) and secondly, the model confidence score C (calculated using the loss function MAE, ranging from 0 to 1, with higher values ​​indicating more reliable predictions).

[0041] The model training data comes from a historical stirring condition database, covering: 1000+ sets of experimental data for different material types (solutions, colloids, pastes, etc.); covering a wide viscosity range of 0-2000 mPa・s, with ≥200 sets of valid data collected for each viscosity range; The actual viscosity was labeled (detected by an offline rotational viscometer and used as the label value y), and the Adam optimizer was used to minimize the mean square error (MSE) between the predicted value and the label value.

[0042] The confidence-driven correction logic is as follows: After the model outputs the results, viscosity correction is performed according to the following rules: If confidence level : Directly use the corrected viscosity value output from the model This serves as the basis for subsequent parameter matching; If confidence level : Triggering collaborative correction of historical data, specifically: Retrieve the current input parameters from the historical operating condition database. The three sets of data with the smallest Euclidean distance; Calculate the average actual viscosity of these three sets of data. According to the formula By fusing model predictions with historical averages, the final corrected viscosity is obtained. This ensures the reliability of the assessment at low confidence levels.

[0043] By leveraging the time-series learning capabilities of LSTM models, the problem of traditional physical formulas failing to adapt to dynamic changes in material properties (such as gradual temperature changes and shear thinning effects) can be solved. For shear-thinning materials (such as polymer solutions), the model can capture the hysteresis relationship between the increase in rotation speed and the nonlinear decrease in viscosity, reducing the viscosity assessment error from ±15% of the physical formula to within ±5%. The confidence-driven correction logic ensures the continuity of parameter matching by compensating with historical data when the model prediction is unstable (such as in the early stage of material phase change), avoiding drastic fluctuations in stirring parameters due to viscosity misjudgment and improving system robustness.

[0044] According to the appendix Figure 2 - Appendix Figure 3 As shown, in the S4 stirring parameter dynamic feedback adjustment step: The torque feedback adjustment logic includes: Level 1 response: Basic speed reduction adjustment, when real-time torque... At that time, a level 1 response is triggered: According to the formula

[0045] in: This is the deceleration adjustment coefficient, ranging from 0.1 to 0.15. The higher the viscosity, the larger the value, making it suitable for high-resistance scenarios. The current stirring speed, in units of ; The new stirring speed after adjustment, in units of ; After the speed reduction is implemented, a 30-second delay is performed for verification; other parameter adjustments are paused to focus on monitoring the torque change trend and avoid oscillations caused by frequent adjustments. If after Level 1 response Enter stable maintenance mode: Maintain current speed and angle parameters, trigger fluctuation monitoring every 2 minutes (collect torque standard deviation). threshold Determined to be stable); like A limited number of further speed reductions will be implemented, with a cumulative speed reduction of [amount missing]. The original value should be used to avoid over-adjustment that could worsen uniformity. Secondary response: Angle coordination adjustment, if after the primary response This triggers a level 2 response; Adjust the blade angle by 5°-10° and simultaneously start a 30-second verification cycle to monitor the torque drop. If the torque is still Perform extreme speed reduction and angle readjustment: reduce the speed to the lowest speed in the viscosity range (e.g., viscosity range 3 corresponds to the lower limit of 800 r / min, which needs to be combined with the specific range parameters); further reduce the angle by 5° and start a 60-second countdown protection to forcibly reserve a safety buffer time; When the countdown ends Triggering limit protection: The system alarms and forces the speed down to the minimum safe speed (fixed value 200r / min, to ensure the transition state before the motor stops). If the torque drops back to a safe range After two consecutive successful verifications, the parameter recovery process was initiated. Speed ​​recovery: Verify the torque by increasing the speed by 5% each time (e.g., 200 r / min → 210 r / min → …) with a 30-second interval. Angle recovery: It is performed in increments of 2° each time (e.g., 40°→42°→…), in coordination with speed recovery, to avoid sudden changes in a single parameter.

[0046] By implementing a layered response of "first-level speed reduction - second-level angle adjustment - limit protection", the torque is controlled in a stepped manner to avoid overload damage to the equipment due to sudden torque increases. At the same time, the 30-second verification mechanism reduces frequent adjustments, improves equipment stability, and lowers the failure rate.

[0047] According to the appendix Figure 2 - Appendix Figure 4 As shown, the uniformity feedback adjustment logic includes: Uniformity calculation formula:

[0048] in: Turbidity deviation refers to the turbidity deviation at different points within the mixing tank, which is collected in real time by a multi-channel sensor. and These represent the maximum and minimum turbidity values ​​monitored during the stirring cycle; when When the time is right, a level one response is triggered; if after two consecutive adjustments... This triggers a level 2 response; First-level response: Differentiated adjustment across viscosity ranges: Low viscosity range adjustment strategies for ranges 1 and 2: Prioritize increasing the rotational speed to enhance fluid disturbance, and perform "increase rotational speed by 10% + 1 minute verification".

[0049] If verified Maintain the current speed for 3 minutes, then drop it back to the upper limit of the range (e.g., the upper limit of 800 r / min corresponds to range 2) to balance energy consumption and uniformity. If verified And the foam is abnormal (the turbidity sensor identifies the foam signal): switch to "angle compensation + speed reduction", that is, increase the angle by 3°-5° and reduce the speed by 5% (suppress the foam while maintaining shear force); High viscosity range adjustment strategy (range 3-5): Enhance material shearing through angle optimization; dynamically calculate the new angle using a formula:

[0050] in: This is the angle adjustment coefficient, ranging from 5° to 10°. The higher the viscosity, the larger the value, making it suitable for high-resistance materials. This is the current blade angle; The blade angle has been adjusted. Execute "adjust to larger angle" "+1 minute verification": like Maintain the current angle parameters to ensure mixing stability; like : Superimpose “angle + speed coordinated adjustment”, that is, further increase the angle by 3°-5° + increase the speed by 5% (strengthen the coupling between shear and fluid motion); Level 2 Response: Coordinated Compensation of Agents and Time After two consecutive Level 1 adjustments This triggers a level 2 response: Chemical dosing: For additional chemical dosing in step S5, the dosage is calculated using the following formula: Increase the dosage of reagents (temporarily increase k to 1.2-1.3 to accelerate material homogenization); Time compensation: Extend the stirring time by 10% to 20%, and implement it simultaneously with the addition of the reagent to ensure thorough mixing; After adjustment, re-verify U; if it rises back to If the parameters fall back to normal, or if they still do not meet the requirements, a cross-correction logic is triggered.

[0051] By employing a segmented adjustment strategy (adjusting speed for low viscosity and adjusting angle for high viscosity) and a two-stage response of reagents and time, the problem of improving the uniformity of materials with different viscosities is solved, resulting in a stable improvement in mixing uniformity.

[0052] According to the appendix Figure 2 - Appendix Figure 5 As shown, the cross-correction logic specifically includes the following hierarchical correction strategies: I. Low viscosity range correction scenario for ranges 1 and 2: Scenario 1: Reducing speed to maintain torque leads to a decrease in uniformity; When speed reduction is implemented due to excessive torque, and the uniformity U decreases by ≥10%, such as from 90% to ≤80%, a correction is triggered. Speed ​​compensation: Increase the speed to 90% to 95% of the speed before deceleration (strictly not exceeding the upper limit of the original range, such as 800 r / min for the original upper limit of range 2, then ≤760 r / min after compensation). Angle coordination: Simultaneously increase the blade angle by 2°-3° (to enhance shear force compensation and improve mixing efficiency); Validate closed loop: After compensation, maintain uniformity monitoring for 1 minute. If Then keep the current parameters; If still This triggers an extension of the mixing time (extending the base mixing time by 10% to compensate for homogenization through time). Scenario 2: Increasing speed to improve uniformity causes torque to approach the threshold (stop speed + angle / time adjustment); due cause After increasing the engine speed, the torque rises to (Not exceeding the limit but close to the threshold), triggering correction: Pause speed adjustment: Immediately stop any further speed increase operations; Angle priority: Increase the blade angle by 3°-5° (to improve uniformity by increasing the material contact area, thus mitigating the torque risk associated with high speed). Time compensation: Simultaneously extend the stirring time by 5% (in conjunction with angle adjustment to ensure mixing effect); Verification cycle: Monitor torque and uniformity every 2 minutes; if torque drops back to... and Restore normal adjustment logic; II. High viscosity range correction scenario in interval 3-5; Scenario 1: Increasing the angle to improve uniformity causes a sudden increase in torque (angle adjustment + slight decrease in engine speed). due cause After adjusting the angle, the torque surged to [a certain value] within 30 seconds. Triggering a fix: Emergency Reversal: The angle is immediately reversed to 90% of its original value (e.g., if the original angle is 45°, it is reversed to 40.5°), while the rotational speed decreases by 3% to 5% (quickly reducing the equipment load). Step-by-step verification: After the pullback, monitor the torque for 1 minute. If it stabilizes... Adjust the angle in steps of 50% of the original adjustment range (each step is 20 seconds apart; for example, if the original plan was to adjust by 5°, adjust by 2.5° each time to avoid torque fluctuations). Reassessment Trigger: If torque remains after callback Forced viscosity reassessment (re-execute the LSTM model and verify viscosity assessment bias).

[0053] Scenario 2: Slowing down to maintain torque results in consistently low uniformity (angle priority + reagent assistance) After two consecutive decelerations, the torque stabilizes at... but Triggering a fix: Angle enhancement: Based on the current rotational speed, according to the formula Increase the angle ( This is the original angle adjustment coefficient, such as the original... =10° then new =5°, reducing the risk of overload in the adjustment range); Step-by-step verification: Monitor uniformity 30 seconds after each angle increase. If U rises back to... Then maintain; if the angle has reached the upper limit of the interval (e.g., the upper limit of interval 5 is 75°), trigger: Chemical adjuvant: Add 3% to 5% of the chemical, calculated according to the following formula: The dosage of the additional drug (k) was temporarily increased to 1.15-1.2. Time compensation: Extend stirring time by 15% (to synergistically improve homogenization effect with the agent).

[0054] Universal extreme scenario correction scenario across the entire range; Scenario 1: Dual failures continue to worsen, torque and ; When neither torque nor uniformity improves after two consecutive adjustments, an extreme correction is triggered. Safe speed reduction: Force speed reduction to 80% of the lowest speed in the range (e.g., if the lowest speed in range 3 is 800 r / min, then reduce to 640 r / min), while adjusting the blade angle to the median of the range (e.g., if the median angle in range 3 is 38°, to balance the load and mixing). Mixed mode: Switches between "low speed high angle (maintained for 60 seconds) → medium speed low angle (maintained for 60 seconds)" every 2 minutes (e.g., interval 3: 640r / min + 38° → 1000r / min + 33°) to break up material agglomeration through alternating flow fields; Manual warning: If the standard is not met after 5 minutes, an audible and visual alarm and a remote notification will be triggered to prompt manual inspection of the material properties, sensor status, or equipment mechanical failure.

[0055] Scenario 2: Parameter adjustment cumulative limit exceeded (speed / angle adjustment range too large) If the cumulative speed adjustment is ≥30% or the cumulative angle adjustment is ≥20° within 10 minutes, a correction will be triggered. Historical comparison: The system automatically retrieves the best parameter combination in the same viscosity range (efficient operating parameters stored in the database). Deviation verification: If the current parameter deviates from the historical best by ≥20%, execute parameter reset: Speed ​​recovery: The speed gradually decreases to the historical benchmark speed in increments of 5% each time (torque is verified at 1-minute intervals). Angle correction: Each time, the angle gradually decreases to the historical reference angle in increments of 2° (in coordination with speed correction to avoid sudden changes in a single parameter). Restart Adjustment: After resetting, restart the normal torque and uniformity parameter adjustment logic to ensure that the parameters are dynamically adapted within the optimal range.

[0056] By cross-correcting across different intervals and scenarios, the coupling contradiction between torque and uniformity adjustment is resolved (such as reducing speed to ensure safety but sacrificing uniformity). Under extreme working conditions, mode switching and parameter resetting ensure system stability, thereby improving the adaptability of mixing complex materials.

[0057] An intelligent stirring control system is used to execute an intelligent stirring control method, comprising: The system parameter configuration and self-test module, with its preset basic thresholds and adjustment parameters, performs functional verification of the sensor and loads a pre-trained LSTM viscosity model. The material import and basic status monitoring module performs the material import operation and simultaneously collects basic status parameters such as initial liquid level and temperature. The viscosity dynamic assessment and initial parameter matching module collects dynamic and static parameters, evaluates the material viscosity in real time through the LSTM model and matches it to the corresponding range, and calls the initial stirring parameters for that range. The stirring parameter dynamic feedback adjustment module adjusts parameters based on real-time feedback of torque and uniformity, and executes cross-correction logic. The reagent dosing module performs the quantitative addition of reagents based on viscosity assessment results and stirring status. The material discharge and status verification module performs the discharge operation, monitors the status parameters of the discharge process, and triggers a new cycle after the discharge is completed.

[0058] By presetting initial parameters in five viscosity ranges and combining them with an LSTM model to correct viscosity assessment in real time, the problem that traditional fixed parameters cannot be adapted to materials with a wide viscosity range of 0-2000 mPa·s is solved. It can achieve precise control for low viscosity solutions, high viscosity pastes, etc., and expand the applicable scenarios.

[0059] The torque feedback mechanism, consisting of a first-level speed reduction, a second-level angle adjustment, and a tiered response mechanism for limit protection, can prevent torque from exceeding limits when stirring high-viscosity materials. The uniformity feedback mechanism, which uses a segmented adjustment strategy (adjusting speed for low viscosity and adjusting angle for high viscosity), steadily improves the uniformity of stirring. At the same time, the 30-second / 1-minute verification cycle reduces parameter oscillations and ensures system stability.

[0060] The cross-correction logic addresses several typical coupled scenarios, such as "reducing speed to ensure safety but sacrificing uniformity" and "adjusting angle to improve uniformity but causing a sudden increase in torque." Through a coordinated strategy of speed compensation, angle fine-tuning, and reagent assistance, it can maintain stable operation even under extreme conditions (such as double non-compliance or parameter cumulative over-limit), thus improving the adaptability to complex material mixing.

[0061] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0062] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent stirring control method, characterized in that, Includes the following steps: S1. System parameter configuration and self-test: preset basic thresholds and adjustment parameters, perform functional verification of the sensor, and load the pre-trained LSTM viscosity model; S2. Material introduction and basic status monitoring: Perform the material introduction operation and simultaneously collect basic status parameters such as initial liquid level and temperature. S3. Viscosity dynamic evaluation and initial parameter matching: Collect dynamic and static parameters, evaluate the material viscosity in real time through the LSTM model and match it to the corresponding range, and call the initial stirring parameters of that range. S4. Dynamic feedback adjustment of stirring parameters: Parameters are adjusted based on real-time feedback of torque and uniformity, and cross-correction logic is executed. S5. Quantitative addition of reagents: Perform quantitative addition of reagents based on viscosity assessment results and stirring status; S6. Material Discharge and Status Verification: Execute the discharge operation, monitor the status parameters during the discharge process, and trigger a new cycle after the discharge is completed.

2. The intelligent stirring control method according to claim 1, characterized in that, In the S1 system parameter configuration and self-test steps: The basic thresholds include upper and lower limits of liquid level, torque safety threshold, uniformity standard, and the rotational speed and blade angle corresponding to each viscosity range; The formula for calculating the torque safety threshold is as follows: ,in: This refers to the motor's rated maximum torque, expressed in N·m. This represents the current material density, in kg / m³. The density is the standard material density, in kg / m³, e.g., 1000 for clean water; The adjustment parameters include the single adjustment range of the rotational speed, the single adjustment range of the blade angle, and the feedback verification interval; Sensor calibration must ensure that the data acquisition frequency of sensors such as torque, turbidity, and temperature is ≥10Hz.

3. The intelligent stirring control method according to claim 2, characterized in that, The rotational speed and blade angle corresponding to each viscosity range are: Range 1 (0-100 mPa・s): Rotation speed 300-500 r / min, blade angle 15-20°; Range 2: 101-500 mPa·s: rotational speed 501-800 r / min, blade angle 21-30°; Range 3: 501-1000 mPa·s; rotational speed 801-1200 r / min; blade angle 31-45°. Range 4: 1001-2000 mPa・s: rotational speed 1201-1500 r / min, blade angle 46-60°; Range 5, above 2001 mPa·s: rotational speed 1501-1800 r / min, blade angle 61-75°.

4. The intelligent stirring control method according to claim 1, characterized in that, In the S2 material introduction and basic condition monitoring step, the basic condition parameters collected include: Initial liquid level height, in meters; Initial temperature of the material, in °C; Import rate, measured in m³ / min, is recorded in real time via a flow sensor.

5. The intelligent stirring control method according to claim 1, characterized in that, In the S3 viscosity dynamic evaluation and initial parameter matching step: The dynamic parameters include real-time torque T, motor current, liquid level fluctuation value and stirring speed n, and the unit of real-time torque T is N·m, the unit of motor current is A, the unit of liquid level fluctuation value is m, and the unit of stirring speed n is r / min. Static parameters include the real-time temperature of the material. and material density , The unit is ℃; The viscosity evaluation steps are as follows: Preliminary estimate: Where K is the equipment characteristic coefficient, and it is taken as 0.5 for propulsion equipment and 0.8 for turbine equipment. For rotational speed, and The unit is d is the blade diameter, and the unit is m; LSTM Model Correction: Input Parameters such as initial temperature and density are used to output the final viscosity value. And the confidence level. If the confidence level is ≥90%, it is adopted directly; if the confidence level is <90%, it is adjusted based on historical data.

6. The intelligent stirring control method according to claim 3, characterized in that, In the S4 dynamic feedback adjustment step of stirring parameters, parameter adjustment is performed based on real-time feedback of torque and uniformity, wherein the torque feedback adjustment logic includes: The torque feedback adjustment logic includes: Level 1 response: Basic speed reduction adjustment, when real-time torque... At that time, a level 1 response is triggered: According to the formula ; in: This is the speed reduction adjustment coefficient, with a value ranging from 0.1 to 0.15; The current stirring speed, in units of ; The new stirring speed after adjustment, in units of ; After the speed reduction is implemented, there will be a 30-second delay for verification. If after Level 1 response Enter stable maintenance mode: Maintain current speed and angle parameters, and trigger fluctuation monitoring every 2 minutes; like A limited number of further speed reductions will be implemented, with a cumulative speed reduction of [amount missing]. Original value; Secondary response: Angle coordination adjustment, if after the primary response This triggers a level 2 response; Adjust the blade angle by 5°-10° and simultaneously start a 30-second verification cycle to monitor the torque drop. If the torque is still Perform extreme speed reduction and angle readjustment: reduce speed to the lowest speed in the viscosity range; further reduce angle by 5° and start 60-second countdown protection to force a safety buffer time. When the countdown ends Triggering limit protection: The system alarms and forcibly reduces the speed to the minimum safe speed; If the torque drops back to a safe range After two consecutive successful verifications, the parameter recovery process was initiated. Speed ​​recovery: The torque is verified by increasing the torque by 5% increments each time, with a 30-second interval. Angle recovery: Increments by 2° each time, and is performed in conjunction with speed recovery.

7. The intelligent stirring control method according to claim 6, characterized in that, In the S4 stirring parameter dynamic feedback adjustment step, the uniformity feedback adjustment logic includes: Uniformity calculation formula: ; in: Turbidity deviation refers to the turbidity deviation at different points within the mixing tank. and These represent the maximum and minimum turbidity values ​​monitored during the stirring cycle; when When the time is right, a level one response is triggered; if after two consecutive adjustments... This triggers a level 2 response; First-level response: Differentiated adjustment across viscosity ranges: Low viscosity range adjustment strategies for ranges 1 and 2: Prioritize increasing the rotational speed to enhance fluid disturbance, and perform "increase rotational speed by 10% + 1 minute verification"; If verified Maintain the current speed for 3 minutes, then drop back to the upper limit of the range to balance energy consumption and uniformity; If verified And the foam is abnormal: switch to "angle compensation + speed reduction", that is, increase the angle by 3°-5° and reduce the speed by 5%; High viscosity range adjustment strategy (range 3-5): Enhance material shearing through angle optimization; dynamically calculate the new angle using a formula: ； in: This is the angle adjustment coefficient, with a value of 5°-10°; This is the current blade angle; The blade angle has been adjusted. Execute "Adjust to larger angle" "+1 minute verification": like Maintain the current angle parameters; like The "angle + speed coordinated adjustment" is superimposed, that is, the angle is increased by 3°-5° and the speed is increased by 5%; Level 2 Response: Coordinated Compensation of Agents and Time After two consecutive Level 1 adjustments This triggers a level 2 response: Addition of pesticide: Addition of pesticide in step S5 of the quantitative pesticide addition process; Time compensation: Extend the stirring time by 10% to 20%, and implement it simultaneously with the addition of the reagent; After adjustment, re-verify U; if it rises back to If the parameters fall back to normal, or if they still do not meet the requirements, a cross-correction logic is triggered.

8. The intelligent stirring control method according to claim 7, characterized in that, The cross-correction logic specifically includes the following hierarchical correction strategies: I. Low viscosity range correction scenario for ranges 1 and 2: Scenario 1: Reducing speed to maintain torque leads to a decrease in uniformity; When speed reduction is implemented due to excessive torque, and the uniformity U decreases by ≥10%, such as from 90% to ≤80%, a correction is triggered. Speed ​​compensation: Increases the speed to 90% to 95% of the speed before deceleration; Angle coordination: Synchronously increase the blade angle by 2°-3°; Validate closed loop: After compensation, maintain uniformity monitoring for 1 minute. If Then keep the current parameters; If still This triggers an extension of the stirring time; Scenario 2: Increasing engine speed and uniformity leads to torque approaching the threshold. due cause After increasing the engine speed, the torque rises to Triggering a fix: Pause speed adjustment: Immediately stop any further speed increase operations; Angle priority: Increase the blade angle by 3°-5°; Time compensation: Simultaneously extend the stirring time by 5%; Verification cycle: Monitor torque and uniformity every 2 minutes; if torque drops back to... and Restore normal adjustment logic; II. High viscosity range correction scenario in interval 3-5; Scenario 1: Increasing the angle to improve uniformity causes a sudden increase in torque; due cause After adjusting the angle, the torque surged within 30 seconds. Triggering a fix: Emergency correction: The angle is immediately corrected to 90% of its original value, while the engine speed decreases by 3% to 5%; Step-by-step verification: After the pullback, monitor the torque for 1 minute. If it stabilizes... Adjust the angle gradually in steps, increasing the adjustment range by 50% of the original range; Reassessment Trigger: If torque remains after callback Forced viscosity reassessment; Scenario 2: Slowing down to maintain torque results in consistently low uniformity; After two consecutive decelerations, the torque stabilizes at... but Triggering a fix: Angle enhancement: Based on the current rotational speed, according to the formula Increase the angle; Step-by-step verification: Monitor uniformity 30 seconds after each angle increase. If U rises back to... Then maintain; If the angle has reached the upper limit of the interval, trigger: Pharmaceutical adjuvant: Add 3% to 5% of the drug; Time compensation: Extend mixing time by 15%; Universal extreme scenario correction scenario across the entire range; Scenario 1: Dual failures continue to worsen, torque and ; When neither torque nor uniformity improves after two consecutive adjustments, an extreme correction is triggered. Safe speed reduction: Force speed reduction to 80% of the lowest speed in the range, while adjusting the blade angle to the median of the range; Hybrid mode: Switches between "low speed high angle → medium speed low angle" every 2 minutes to break up material agglomeration through alternating flow fields; Manual warning: If the standard is not met after 5 minutes, an audible and visual alarm and a remote notification will be triggered; Scenario 2: Parameter adjustment cumulatively exceeds limits; If the cumulative speed adjustment is ≥30% or the cumulative angle adjustment is ≥20° within 10 minutes, a correction will be triggered. Historical comparison: The system automatically retrieves the best parameter combination in history for the same viscosity range; Deviation verification: If the current parameter deviates from the historical best by ≥20%, execute parameter reset: Speed ​​correction: The speed gradually decreases back to the historical baseline speed in increments of 5% each time; Angle correction: Each time, the angle gradually decreases to the historical baseline angle in increments of 2°; Restart Adjustment: After resetting, restart the normal torque and uniformity parameter adjustment logic.

9. An intelligent stirring control system, used to execute the intelligent stirring control method described in any one of 1-8, characterized in that, include: The system parameter configuration and self-test module, with its preset basic thresholds and adjustment parameters, performs functional verification of the sensor and loads a pre-trained LSTM viscosity model. The material import and basic status monitoring module performs the material import operation and simultaneously collects basic status parameters such as initial liquid level and temperature. The viscosity dynamic assessment and initial parameter matching module collects dynamic and static parameters, evaluates the material viscosity in real time through the LSTM model and matches it to the corresponding range, and calls the initial stirring parameters for that range. The stirring parameter dynamic feedback adjustment module adjusts parameters based on real-time feedback of torque and uniformity, and executes cross-correction logic. The reagent dosing module performs the quantitative addition of reagents based on viscosity assessment results and stirring status. The material discharge and status verification module performs the discharge operation, monitors the status parameters of the discharge process, and triggers a new cycle after the discharge is completed.

Images