Slurry stirring control method, system and device
By collecting motor power signals and slurry status information in real time and using machine learning models to dynamically adjust the slurry mixing control strategy, the problems of low mixing efficiency and unstable effect in the existing technology are solved, and efficient slurry mixing control is achieved.
Patent Information
- Application Number
- CN202511567463.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-30
- Publication Date
- 2026-01-27
AI Technical Summary
In the current slurry mixing process, the mixing control parameters cannot be adjusted in real time, resulting in low mixing efficiency and inability to ensure mixing effect. In particular, when there are differences in material batches and changes in ambient temperature, the quality of the slurry cannot be guaranteed.
By collecting the motor power signal and slurry status information of the mixing equipment in real time, the machine learning model is used to determine the real-time status type of the slurry, and the mixing control strategy, including speed, running time and temperature control, is dynamically adjusted according to different status types.
It enables real-time, dynamic adjustment of control strategies during the mixing process, improving mixing efficiency and ensuring mixing effect, while adapting to batch differences in materials and environmental changes.
Smart Images
Figure CN121401940A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of slurry mixing technology, and more particularly to a method, system, and apparatus for controlling slurry mixing. Background Technology In all technological fields requiring high-quality slurry mixing, such as batteries, supercapacitors, coatings, and inks, the control of the slurry mixing process is particularly important. For example, the dispersion uniformity, stability, and conductive network integrity of the electrode slurry directly determine the thickness consistency and areal density uniformity of the electrode sheet, which in turn affects the core performance of the battery, such as energy density, cycle life, safety, and self-discharge rate.
[0002] Slurry mixing is a complex physicochemical coupling process that requires five stages: powder wetting, agglomerate breaking, binder coating, uniform dispersion of components, and system maturation and stabilization. Currently, the quality control of slurry mixing still relies mainly on "experience-based + offline lag detection": engineers set mixing time and speed curves based on experience, take samples after mixing, and judge the quality through subsequent testing; if it is unqualified, the entire batch of slurry must be scrapped or reworked. However, due to differences in material batches and possible changes in environmental temperature, this method uses a fixed program control strategy, which cannot adjust the mixing control parameters in real time, resulting in low mixing efficiency and an inability to ensure the mixing effect. Summary of the Invention
[0003] This invention provides a method for controlling slurry mixing, which dynamically adjusts the mixing control strategy to improve mixing efficiency while ensuring mixing effect. The method includes: During the slurry mixing process, the motor power signal of the mixing equipment and the slurry status information are acquired within a preset time period; the slurry status information includes the slurry temperature information, viscosity information and conductivity information; The motor power signal and slurry status information are input into a preset slurry status classification model to determine the real-time status type of the slurry. The slurry status classification model is obtained by training a machine learning model using multiple sets of historical data. Each set of historical data includes historical motor power signals, historical status information, and the corresponding historical status type. Based on the real-time state type of the slurry and the pre-set stirring control methods corresponding to different state types, the target stirring control method is determined. The target stirring control method is the stirring control method corresponding to the real-time state type. Based on the target stirring control method, motor power signal and slurry state information, the stirring control strategy is determined.
[0004] Optionally, before inputting the motor power signal and slurry status information into a preset slurry status classification model to determine the real-time status type of the slurry, the following steps are also included: Feature extraction was performed on the motor power signal and the slurry state information to obtain power spectrum features, viscosity change rate features, conductivity area change features and temperature features; The motor power signal and slurry status information are input into a preset slurry status classification model to determine the real-time status type of the slurry, including: By inputting the power spectrum characteristics, viscosity change rate characteristics, conductivity area change characteristics, and temperature characteristics into a preset slurry state classification model, the real-time state type of the slurry is determined.
[0005] Optionally, the real-time state types of the slurry include: normal dispersion period, agglomeration risk period, conductive network formation period, and mixing endpoint.
[0006] Optionally, a mixing control strategy is determined based on the target mixing control method, motor power signal, and slurry status information, including: When the real-time state of the slurry is in the normal dispersion period, the stirring control strategy is determined by analyzing the motor power signal and the state information of the slurry based on the model predictive control algorithm. When the real-time state of the slurry is in the agglomeration risk period, the motor power signal and the state information of the slurry are analyzed according to the preset rule database and reinforcement learning algorithm to determine the stirring control strategy. When the real-time state of the slurry is in the conductive network formation stage, the stirring control strategy is determined by analyzing the motor power signal and the state information of the slurry based on the fuzzy logic control algorithm. When the real-time state type of the slurry is the mixing endpoint, the mixing control strategy is determined according to the pre-set mixing endpoint control strategy.
[0007] Optionally, the stirring control strategy may include at least one of the stirring equipment speed control strategy, runtime control strategy, and temperature control strategy.
[0008] Optionally, the slurry state information may also include particle size information, which includes particle size distribution curves and concentration information.
[0009] Optionally, the structure of the slurry state classification model can be any one of a temporal convolutional network (TCN), a long short-term memory network (LSTM), or a neural network based on a self-attention mechanism.
[0010] This invention also provides a slurry mixing control system for dynamically adjusting the mixing control strategy to improve mixing efficiency while ensuring mixing effect. The slurry mixing control system includes: an online viscometer, an online conductivity meter, a temperature sensor, a motor power sensor, and a control device. An online viscometer is configured to collect viscosity information of the slurry; An online conductivity meter is configured to collect conductivity information of the slurry; A temperature sensor is configured to collect temperature information of the slurry. A motor power sensor is configured to collect the motor power signal of the mixing equipment; The control device is configured to implement the control method for mixing the slurry as described above.
[0011] This invention also provides a slurry mixing control device for dynamically adjusting the mixing control strategy to improve mixing efficiency while ensuring mixing effect. The slurry mixing control device includes: The data acquisition module is used to acquire the motor power signal of the mixing equipment and the state information of the slurry within a preset time period during the slurry mixing process; the state information of the slurry includes the temperature information, viscosity information and conductivity information of the slurry. The classification module is used to input the motor power signal and the slurry status information into a preset slurry status classification model to determine the real-time status type of the slurry. The slurry status classification model is obtained by training a machine learning model using multiple sets of historical data. Each set of historical data includes historical motor power signals, historical status information, and the corresponding historical status type. The strategy determination module is used to determine the target mixing control method based on the real-time state type of the slurry and the pre-set mixing control methods corresponding to different state types. The target mixing control method is the mixing control method corresponding to the real-time state type. The mixing control strategy is determined based on the target mixing control method, the motor power signal and the state information of the slurry.
[0012] This invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the above-described slurry stirring control method.
[0013] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described slurry stirring control method.
[0014] This invention also provides a computer program product, which includes a computer program that, when executed by a processor, implements the above-described slurry stirring control method.
[0015] In this embodiment of the invention, during slurry mixing, the motor power signal of the mixing equipment and the state information of the slurry are acquired within a preset time period. The motor power signal and the slurry state information are input into a preset slurry state classification model to determine the real-time state type of the slurry. Based on the real-time state type of the slurry and the pre-set mixing control methods corresponding to different state types, a target mixing control method is determined. Finally, based on the target mixing control method, the motor power signal, and the slurry state information, a mixing control strategy is determined. Thus, compared to existing fixed-program control strategies, by acquiring the motor power signal and slurry state information in real time, using the slurry state classification model to determine the real-time state type of the slurry, and determining the target mixing control method based on the pre-set mixing control methods corresponding to different state types, and then determining the mixing control strategy based on the target mixing control method, real-time and dynamic adjustment of the mixing control strategy can be achieved, improving mixing efficiency while ensuring mixing effect. Attached Figure Description
[0016] 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.
[0017] In the attached diagram: Figure 1 This is a structural diagram of a slurry mixing control system provided in an embodiment of the present invention; Figure 2 A flowchart of a slurry stirring control method provided in an embodiment of the present invention; Figure 3 A flowchart of another slurry stirring control method provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of a slurry mixing control device provided in an embodiment of the present invention; Figure 5 This is a schematic diagram of a computer device provided in an embodiment of the present invention. Detailed Implementation
[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings. Here, the illustrative embodiments of the present invention and their descriptions are used to explain the present invention, but are not intended to limit the present invention.
[0019] In the description of this specification, the terms "comprising," "including," "having," and "containing" are open-ended terms, meaning that they include but are not limited to. The terms "an embodiment," "a specific embodiment," "some embodiments," and "for example," etc., refer to specific features, structures, or characteristics described in connection with that embodiment or example that are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, or characteristics described can be combined in any suitable manner in one or more embodiments or examples. The order of steps involved in the various embodiments is used to illustrate the implementation of this application, and the order of steps is not limited and can be adjusted appropriately as needed.
[0020] In the description of this specification, the terms "first" and "second," etc., are used to distinguish different objects or to distinguish different treatments of the same object, rather than to describe a specific order of objects.
[0021] In the description of this specification, "and / or" is merely a way of describing the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone.
[0022] Research has found that existing slurry mixing methods typically rely on engineers' experience to set fixed program control strategies. However, due to batch variations in materials and changes in ambient temperature and humidity, this method cannot adjust the mixing process in real time, thus failing to ensure that the slurry quality reaches its optimal state.
[0023] Based on this, embodiments of the present invention provide a control scheme for slurry mixing, which can adjust the mixing strategy in real time to ensure that the slurry quality reaches the optimal state.
[0024] Figure 1 This is a structural diagram of a slurry mixing control system provided in an embodiment of the present invention. Figure 1 As shown, the control system for slurry mixing may include an online viscometer 1, an online conductivity meter 2, a temperature sensor 3, a motor power sensor 4, and a control device 5; Online viscometer 1 is configured to collect viscosity information of the slurry; Online conductivity meter 2 is configured to collect conductivity information of slurry; Temperature sensor 3 is configured to collect temperature information of the slurry; Motor power sensor 4 is configured to collect the motor power signal of the mixing equipment; Control device 5 is configured to implement the following control method for mixing slurry.
[0025] In practical implementation, an online viscometer 1, an online conductivity meter 2, a temperature sensor 3, and a motor power sensor 4 can be installed on the mixing equipment. The online viscometer 1, based on the principle of rotation or vibration, collects the viscosity information of the slurry in real time; the online conductivity meter 2 uses a four-electrode method to measure the conductivity information of the slurry in real time, and this conductivity information can be used to determine the dispersion state of the conductive agent; the temperature sensor 3 can collect the temperature information of the slurry in real time; and the motor power sensor 4 can indirectly determine the consistency and dispersion resistance of the slurry by collecting the motor power signal of the mixing equipment in real time.
[0026] The aforementioned online viscometer 1, online conductivity meter 2, temperature sensor 3, and motor power sensor 4 can synchronously acquire data and perform analog-to-digital conversion via a high-speed data acquisition card, and then transmit the data to the control device 5. The control device 5 processes the acquired data (motor power signal, slurry temperature information, viscosity information, and conductivity information) to determine the stirring control strategy. Then, it sends the stirring control strategy to the stirring equipment to control the stirring equipment to execute the stirring control strategy and provides feedback on the execution results.
[0027] It should be noted that the stirring equipment belongs to the execution layer, including the actuator and the PLC controller 6. The aforementioned online viscometer 1, online conductivity meter 2, temperature sensor 3, and motor power sensor 4 are installed in the actuator of the stirring equipment. The PLC controller 6 communicates with the control device 5 to receive the stirring control strategy, control the actuator to execute control commands, and obtain the execution results of the online viscometer 1, online conductivity meter 2, temperature sensor 3, and motor power sensor 4 of the actuator and feed them back to the control device 5 to ensure the real-time performance and synchronization of the control loop.
[0028] It should also be noted that the slurry state information can also include particle size information, which includes particle size distribution curves and concentration information. Therefore, the aforementioned slurry mixing control system can also include an online particle size analyzer (using the principle of laser diffraction) to acquire particle size information, providing more input features for the slurry state classification model and further improving classification accuracy.
[0029] Figure 2 This is a flowchart of a slurry mixing control method provided in an embodiment of the present invention. The method is applied to... Figure 1 The control system for mixing the slurry is shown. For example... Figure 2 As shown, the method for controlling the mixing of the slurry may include: Step 201: During the slurry mixing process, acquire the motor power signal of the mixing equipment and the slurry status information within a preset time period; the slurry status information includes the slurry temperature information, viscosity information and conductivity information. Step 202: Input the motor power signal and slurry status information into a preset slurry status classification model to determine the real-time status type of the slurry; the slurry status classification model is obtained by training a machine learning model using multiple sets of historical data; each set of historical data includes historical motor power signals, historical status information and corresponding historical status types; Step 203: Determine the target mixing control method based on the real-time state type of the slurry and the pre-set mixing control methods corresponding to different state types. The target mixing control method is the mixing control method corresponding to the real-time state type. Determine the mixing control strategy based on the target mixing control method, motor power signal and slurry state information.
[0030] In this embodiment of the invention, during slurry mixing, the motor power signal of the mixing equipment and the state information of the slurry are acquired within a preset time period. The motor power signal and the slurry state information are input into a preset slurry state classification model to determine the real-time state type of the slurry. Based on the real-time state type of the slurry and the pre-set mixing control methods corresponding to different state types, a target mixing control method is determined. Finally, based on the target mixing control method, the motor power signal, and the slurry state information, a mixing control strategy is determined. Thus, compared to existing fixed-program control strategies, by acquiring the motor power signal and slurry state information in real time, using the slurry state classification model to determine the real-time state type of the slurry, and determining the target mixing control method based on the pre-set mixing control methods corresponding to different state types, and then determining the mixing control strategy based on the target mixing control method, real-time and dynamic adjustment of the mixing control strategy can be achieved, improving mixing efficiency while ensuring mixing effect.
[0031] The control method for the above-mentioned slurry mixing is explained below.
[0032] In step 201 above, it can be done by Figure 1 The system includes an online viscometer 1, an online conductivity meter 2, a temperature sensor 3, and a motor power sensor 4 to acquire motor power signals and slurry status information.
[0033] In practice, time-series data within a preset time period is acquired in real time. The preset time period can refer to the time within a preset duration from the current moment to the moment before the current moment. For example, the motor power signal of the mixing equipment and the status information of the slurry are acquired at the current moment and in the past 10 hours.
[0034] It should be noted that after obtaining the motor power signal of the mixing equipment and the state information of the slurry within the preset time period, the original motor power signal and slurry state information can be preprocessed. The preprocessing can include using Z-score normalization based on a sliding window to eliminate dimensional differences, and combining moving average filtering and wavelet transform for joint denoising, so as to effectively suppress interference while retaining the true signal characteristics.
[0035] In step 202 above, the motor power signal and the state information of the slurry can be input into the slurry state classification model according to the pre-trained slurry state classification model to determine the real-time state type of the slurry.
[0036] In practice, multiple sets of historical data can be used to train a machine learning model to obtain a slurry state classification model. The machine learning model can be a model capable of processing time series data. The historical data can be obtained from sources such as historical production data, historical experimental data, and historical simulation data to ensure that the trained model has both experience learning and anomaly diagnosis capabilities, and possesses good accuracy and robustness.
[0037] Specifically, a large amount of historical data can be divided into training and testing sets according to a preset ratio. The training set is used to train a machine learning model to obtain a slurry state classification model, and the testing set is used to test the trained slurry state classification model. Then, the real-time state type of the slurry is determined using the trained slurry state classification model.
[0038] In one embodiment, the structure of the slurry state classification model (machine learning model) can be any of the following: Temporal Convolutional Network (TCN), Long Short-Term Memory Network (LSTM), or Neural Network based on self-attention mechanism.
[0039] For example, assuming the slurry state classification model has a Temporal Convolutional Network (TCN) structure, the motor power signal and slurry state information are input into the trained TCN model, and the TCN model outputs the real-time state type of the slurry. In this embodiment of the invention, the TCN model adopts an expanded causal convolution and residual connection structure, which can exponentially expand its receptive field, effectively capturing long-term causal relationships. Furthermore, the TCN network structure is more stable, training is faster, and its forward structure avoids the gradient vanishing problem, making it easier to train a reliable model from industrial data. Therefore, using the TCN model to classify and predict the state type of slurry yields more accurate results.
[0040] In one embodiment, such as Figure 3 As shown, before step 202 and after step 201, the above-mentioned slurry mixing control method may further include step 301: Feature extraction was performed on the motor power signal and the slurry state information to obtain power spectrum features, viscosity change rate features, conductivity area change features and temperature features; Step 202 may specifically include: By inputting the power spectrum characteristics, viscosity change rate characteristics, conductivity area change characteristics, and temperature characteristics into a preset slurry state classification model, the real-time state type of the slurry is determined.
[0041] In practical implementation, to further improve the classification accuracy of the slurry state classification model, feature extraction can be performed on the motor power signal and the slurry state information to obtain key feature vectors that reflect changes in the slurry state. Specifically, envelope spectrum analysis can be used to extract power spectrum features reflecting particle collisions and aggregation from the motor power signal; viscosity change rate features from the viscosity information; conductivity area change features reflecting the formation of conductive networks from the conductivity information; and temperature features from the temperature information. Then, the power spectrum features, viscosity change rate features, conductivity area change features, and temperature features are input into the slurry state classification model to determine the real-time state type of the slurry.
[0042] In practice, in step 301 above, features can also be extracted using a one-dimensional convolutional neural network, further improving the accuracy of feature extraction. The one-dimensional convolutional neural network can be implemented using existing technologies, and will not be elaborated upon here.
[0043] It should be noted that feature extraction may also include extracting ripple coefficient features from the motor power signal, thereby indirectly evaluating the uniformity of particle size distribution in the slurry. In this embodiment of the invention, when training the slurry state classification model using historical data, feature extraction of the historical data can also be performed using the same method as in step 301 above, and the extracted historical feature data can be used to train the slurry state classification model.
[0044] In step 203 above, a target mixing control method can be determined based on the real-time state type of the slurry and the mixing control methods corresponding to different state types that are preset. The target mixing control method is the mixing control method corresponding to the real-time state type. Then, a mixing control strategy is determined based on the target mixing control method, the motor power signal and the state information of the slurry.
[0045] In practice, different stirring control methods can be set for different states of the slurry. Thus, after obtaining the real-time state of the slurry, the stirring control method corresponding to the real-time state can be determined as the target stirring control method; subsequently, stirring control can be performed according to the target stirring control method, thereby improving the accuracy and efficiency of stirring control.
[0046] In this embodiment of the invention, the real-time state type of the slurry may include: normal dispersion period, agglomeration risk period, conductive network formation period, and stirring endpoint.
[0047] In this embodiment of the invention, the stirring control strategy may include at least one of the stirring equipment speed control strategy, running time control strategy, and temperature control strategy.
[0048] In one embodiment, in step 203 above, when the real-time state type of the slurry is in the normal dispersion period, the motor power signal and the state information of the slurry are analyzed based on the model predictive control algorithm to determine the stirring control strategy.
[0049] In practice, the normal dispersion period may include a stage for crushing large particles. To improve the mixing efficiency of the slurry during the normal dispersion period and ensure the mixing effect, a model predictive control algorithm (MPC algorithm) can be used to analyze the real-time motor power signal and the state information of the slurry, and adjust the mixing control strategy in real time.
[0050] Specifically, the MPC algorithm is a dynamic model that aims to approach the optimal dispersion endpoint as quickly as possible while satisfying constraints such as viscosity / conductivity ratio, temperature limits, and no agglomeration. It adjusts the stirring speed, runtime, and temperature of the agitator in real time. The optimal dispersion endpoint can refer to the slurry's viscosity and conductivity reaching preset values. The current slurry temperature, viscosity, conductivity, and the motor power signal of the agitator are used as state variables for the MPC algorithm, while the stirring speed, runtime, and temperature are used as action control variables to optimize the control trajectory of the stirring equipment's speed, runtime, and temperature over a future period.
[0051] In this way, during the normal dispersion period, the mixing control of the slurry through the MPC algorithm can dynamically adjust the action of the mixing equipment based on the real-time status of the slurry, thereby improving the mixing efficiency while ensuring the mixing effect.
[0052] In one embodiment, in step 203 above, when the real-time state type of the slurry is in the agglomeration risk period, the motor power signal and the state information of the slurry are analyzed according to a preset rule database and reinforcement learning algorithm to determine the stirring control strategy.
[0053] In practice, during the agglomeration risk period, the control objective immediately switches from "efficiency" to "safety and recovery." Specifically, the rule database is a set of "if-then" rules built upon engineers' long-term experience, resulting in extremely fast response times. For example, a rule could be: "If viscosity suddenly increases (real-time viscosity change rate > preset threshold 1) and conductivity suddenly decreases (conductivity area change < preset threshold 2), then immediately execute a rate-reduction cooling emergency operation (control command)." Therefore, during the agglomeration risk period, any abnormality in the motor power signal or the slurry's status information may trigger the rule database's emergency control strategy. For instance, a sudden increase in viscosity and a sudden decrease in conductivity would generate a control command to immediately execute a rate-reduction cooling emergency operation.
[0054] Then, after triggering the emergency control strategy in the rule database, the real-time motor power signal and slurry status information are analyzed based on reinforcement learning (RL algorithm) to generate a recovery control strategy.
[0055] Specifically, after emergency control, the RL algorithm adopts a near-end strategy optimization algorithm to explore fine-tuning control strategies within the safety boundaries set by the rule database (such as rotation speed ±10%). The reward function comprehensively considers the time to reach the optimal dispersion state, the quality score of slurry state information, and energy consumption cost, so as to enable the slurry mixing state to quickly recover to the normal dispersion period.
[0056] In this way, risk control based on the rule database during the agglomeration risk period can quickly suppress the deterioration of the risk, buy time and create initial conditions for subsequent recovery strategies based on reinforcement learning algorithms; then, by exploring recovery control strategies under disturbances through reinforcement learning algorithms, the slurry state can be quickly restored to the normal dispersion period, thereby improving mixing efficiency and ensuring mixing effect.
[0057] In one embodiment, when the real-time state type of the slurry is the conductive network formation period, the stirring control strategy is determined by analyzing the motor power signal and the state information of the slurry according to the fuzzy logic control algorithm.
[0058] In practical implementation, the focus during the conductive network formation stage is on structural protection and optimization. Therefore, a fuzzy logic control algorithm is used, taking the real-time status information of the slurry and the motor power signal as inputs, to determine the rotational speed, running time, and temperature of the mixing equipment. Specifically, the implementation of the fuzzy logic control algorithm can refer to existing technologies, and will not be elaborated further here.
[0059] In one embodiment, when the real-time state type of the slurry is the mixing endpoint, a pre-set mixing endpoint control strategy can be obtained. For example, when the real-time state type of the slurry is the mixing endpoint, the control strategy could be to immediately stop mixing or switch to a minimum energy consumption maintenance mode to prevent "over-mixing".
[0060] In this embodiment of the invention, a control signal is generated based on the above-mentioned stirring control strategy and sent to the stirring equipment. The stirring equipment receives and executes the command, and at the same time, continuously monitors the feedback signal to achieve rolling optimization and closed-loop control.
[0061] Furthermore, after each batch of production (slurry mixing) is completed, the entire process data and slurry quality results are stored in the database for periodic incremental learning of the slurry state classification model, enabling the slurry state classification model to continuously self-optimize and adapt to process changes.
[0062] In summary, the slurry mixing control method provided in this embodiment of the invention firstly collects the time series data of the motor power signal of the mixing equipment and the state information of the slurry in real time, then uses a slurry state classification model to determine the real-time state type of the slurry, and determines the mixing control method corresponding to the real-time state type according to the pre-set mixing control methods corresponding to different state types, and finally determines the mixing control strategy according to the mixing control method corresponding to the real-time state type. This can realize the dynamic adjustment of the mixing control strategy during the mixing process, improving mixing efficiency while ensuring mixing effect.
[0063] This invention also provides a slurry stirring control device, as described in the following embodiments. Since the principle by which this device solves the problem is similar to the slurry stirring control method described above, the implementation of this device can be found in the implementation of the slurry stirring control method, and repeated details will not be elaborated further.
[0064] like Figure 4 The diagram shown is a schematic of a slurry mixing control device provided in an embodiment of the present invention. This control device can be applied to... Figure 1 The control system for mixing the slurry shown may include the following modules: The data acquisition module 401 is used to acquire the motor power signal of the mixing equipment and the state information of the slurry within a preset time period during the slurry mixing process; the state information of the slurry includes the temperature information, viscosity information and conductivity information of the slurry. The classification module 402 is used to input the motor power signal and the slurry status information into a preset slurry status classification model to determine the real-time status type of the slurry. The slurry status classification model is obtained by training a machine learning model using multiple sets of historical data. Each set of historical data includes historical motor power signals, historical status information and corresponding historical status types. The strategy determination module 403 is used to determine the target mixing control mode based on the real-time state type of the slurry and the mixing control mode corresponding to different state types preset in advance. The target mixing control mode is the mixing control mode corresponding to the real-time state type. The mixing control strategy is determined based on the target mixing control mode, the motor power signal and the state information of the slurry.
[0065] In one embodiment, a feature extraction module may also be included, used before the classification module 402 inputs the motor power signal and slurry state information into a preset slurry state classification model to determine the real-time state type of the slurry: Feature extraction was performed on the motor power signal and the slurry state information to obtain power spectrum features, viscosity change rate features, conductivity area change features and temperature features; Classification module 402 is also used for: By inputting the power spectrum characteristics, viscosity change rate characteristics, conductivity area change characteristics, and temperature characteristics into a preset slurry state classification model, the real-time state type of the slurry is determined.
[0066] In one embodiment, the real-time state type of the slurry may include: normal dispersion period, agglomeration risk period, conductive network formation period, and stirring endpoint.
[0067] In one embodiment, the strategy determination module 403 may be specifically used for: When the real-time state of the slurry is in the normal dispersion period, the stirring control strategy is determined by analyzing the motor power signal and the state information of the slurry based on the model predictive control algorithm. When the real-time state of the slurry is in the agglomeration risk period, the motor power signal and the state information of the slurry are analyzed according to the preset rule database and reinforcement learning algorithm to determine the stirring control strategy. When the real-time state of the slurry is in the conductive network formation stage, the stirring control strategy is determined by analyzing the motor power signal and the state information of the slurry based on the fuzzy logic control algorithm. When the real-time state type of the slurry is the mixing endpoint, the mixing control strategy is determined according to the pre-set mixing endpoint control strategy.
[0068] In one embodiment, the stirring control strategy may include at least one of the stirring equipment speed control strategy, runtime control strategy, and temperature control strategy.
[0069] In one embodiment, the structure of the slurry state classification model can be any of the following: Temporal Convolutional Network (TCN), Long Short-Term Memory Network (LSTM), or Neural Network based on Self-Attention Mechanism.
[0070] This invention also provides a computer device. Figure 5This is a schematic diagram of a computer device in an embodiment of the present invention. The computer device 500 includes a memory 510, a processor 520, and a computer program 530 stored in the memory 510 and executable on the processor 520. When the processor 520 executes the computer program 530, it implements the above-mentioned slurry stirring control method.
[0071] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described slurry stirring control method.
[0072] This invention also provides a computer program product, which includes a computer program that, when executed by a processor, implements the above-described slurry stirring control method.
[0073] In this embodiment of the invention, during slurry mixing, the motor power signal of the mixing equipment and the state information of the slurry are acquired within a preset time period. The motor power signal and the slurry state information are input into a preset slurry state classification model to determine the real-time state type of the slurry. Based on the real-time state type of the slurry and the pre-set mixing control methods corresponding to different state types, a target mixing control method is determined. Finally, based on the target mixing control method, the motor power signal, and the slurry state information, a mixing control strategy is determined. Thus, compared to existing fixed-program control strategies, by acquiring the motor power signal and slurry state information in real time, using the slurry state classification model to determine the real-time state type of the slurry, and determining the target mixing control method based on the pre-set mixing control methods corresponding to different state types, and then determining the mixing control strategy based on the target mixing control method, real-time and dynamic adjustment of the mixing control strategy can be achieved, improving mixing efficiency while ensuring mixing effect.
[0074] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0075] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0076] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.
[0077] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.
[0078] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.
Claims
1. A method for controlling slurry mixing, characterized in that, include: During the slurry mixing process, the motor power signal of the mixing equipment and the status information of the slurry are acquired within a preset time period. The state information of the slurry includes its temperature, viscosity, and electrical conductivity. The motor power signal and slurry status information are input into a preset slurry status classification model to determine the real-time status type of the slurry. The slurry status classification model is obtained by training a machine learning model using multiple sets of historical data. Each set of historical data includes historical motor power signals, historical status information, and the corresponding historical status type. Based on the real-time state type of the slurry and the pre-set stirring control methods corresponding to different state types, the target stirring control method is determined. The target stirring control method is the stirring control method corresponding to the real-time state type. Based on the target stirring control method, motor power signal and slurry state information, the stirring control strategy is determined.
2. The method for controlling slurry mixing as described in claim 1, characterized in that, Before inputting the motor power signal and slurry status information into a preset slurry status classification model to determine the real-time status type of the slurry, the following steps are also included: Feature extraction was performed on the motor power signal and the slurry state information to obtain power spectrum features, viscosity change rate features, conductivity area change features and temperature features; The motor power signal and slurry status information are input into a preset slurry status classification model to determine the real-time status type of the slurry, including: By inputting the power spectrum characteristics, viscosity change rate characteristics, conductivity area change characteristics, and temperature characteristics into a preset slurry state classification model, the real-time state type of the slurry is determined.
3. The method for controlling slurry mixing as described in claim 1, characterized in that, The real-time state types of the slurry include: normal dispersion period, agglomeration risk period, conductive network formation period, and mixing endpoint.
4. The method for controlling slurry mixing as described in claim 3, characterized in that, Based on the target mixing control method, motor power signal, and slurry status information, the mixing control strategy is determined, including: When the real-time state of the slurry is in the normal dispersion period, the stirring control strategy is determined by analyzing the motor power signal and the state information of the slurry based on the model predictive control algorithm. When the real-time state of the slurry is in the agglomeration risk period, the motor power signal and the state information of the slurry are analyzed according to the preset rule database and reinforcement learning algorithm to determine the stirring control strategy. When the real-time state of the slurry is in the conductive network formation stage, the stirring control strategy is determined by analyzing the motor power signal and the state information of the slurry based on the fuzzy logic control algorithm. When the real-time state type of the slurry is the mixing endpoint, the mixing control strategy is determined according to the pre-set mixing endpoint control strategy.
5. The method for controlling slurry mixing as described in any one of claims 1-4, characterized in that, The stirring control strategy may include at least one of the following: a stirring speed control strategy, a running time control strategy, and a temperature control strategy.
6. The method for controlling slurry mixing as described in any one of claims 1-4, characterized in that, The state information of the slurry also includes particle size information, which includes particle size distribution curves and concentration information.
7. The method for controlling slurry mixing as described in claim 1, characterized in that, The slurry state classification model can be any one of the following: Temporal Convolutional Network (TCN), Long Short-Term Memory Network (LSTM), or Neural Network based on Self-Attention Mechanism.
8. A control system for slurry mixing, characterized in that, include: Online viscometers, online conductivity meters, temperature sensors, motor power sensors, and control devices; An online viscometer is configured to collect viscosity information of the slurry; An online conductivity meter is configured to collect conductivity information of the slurry; A temperature sensor is configured to collect temperature information of the slurry. A motor power sensor is configured to collect the motor power signal of the mixing equipment; The control device is configured to implement the control method for slurry mixing as described in any one of claims 1-7.
9. A control device for slurry mixing, characterized in that, include: The data acquisition module is used to acquire the motor power signal of the mixing equipment and the status information of the slurry within a preset time period during the slurry mixing process. The state information of the slurry includes its temperature, viscosity, and electrical conductivity. The classification module is used to input the motor power signal and the slurry status information into a preset slurry status classification model to determine the real-time status type of the slurry. The slurry status classification model is obtained by training a machine learning model using multiple sets of historical data. Each set of historical data includes historical motor power signals, historical status information, and the corresponding historical status type. The strategy determination module is used to determine the target mixing control method based on the real-time state type of the slurry and the pre-set mixing control methods corresponding to different state types. The target mixing control method is the mixing control method corresponding to the real-time state type. The mixing control strategy is determined based on the target mixing control method, the motor power signal and the state information of the slurry.
10. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the slurry stirring control method according to any one of claims 1 to 7.