Chemical pulping cooking and washing control method based on ACE technology and storage medium

Abstract

The invention relates to a chemical pulping, cooking and washing control method based on an ACE technology and a storage medium, the method comprises the following steps: collecting process parameter data of a chemical pulping, cooking and washing process, the process parameter data comprising cooking temperature, cooking pressure, pH value, cooking time, liquid ratio, additive concentration, steam flow and liquid level; preprocessing the collected process parameter data, wherein the preprocessing comprises data screening and data filtering; identifying time delay among the process parameter data in the chemical pulping, cooking and washing process; constructing a high-precision multivariable model, and capturing strong nonlinearity and coupling among process parameter data in the chemical pulping, cooking and washing process; and generating an optimal adjustment strategy of the process parameter data of the required target through the multivariable prediction control model. And the influence of single variable adjustment on other variables is avoided, so that the large loss of the production process in the aspects of yield, quality and energy consumption is reduced.

Description

Technical Field

[0001] This application relates to the technical field of chemical pulping, and specifically relates to a chemical pulping cooking and washing control method and storage medium based on the ACE process. Background Art

[0002] The chemical pulping cooking process is an important process in the paper industry, mainly used to separate cellulose from plant raw materials (such as wood, grasses, etc.) and remove lignin and other non-cellulose components. Existing chemical pulping projects are equipped with a DCS control system, and the control loop uses conventional PID control. Due to the process characteristics of long delay, strong coupling, and non-linearity in the pulping cooking and washing process, traditional PID control can only perform single-loop control and it is difficult to achieve multi-variable fine-tuning collaborative online optimization control. During the control process, only the adjustment of a single variable is considered, ignoring the impact of the adjustment of this variable on other variable parameters. For example, the variable coupling in the cooking and washing process may include the mutual influence of temperature and pressure, resulting in large losses in production process in terms of output, quality, and energy consumption. Moreover, PID control has data delay, cannot better achieve the closed-loop control of PID, and single-loop PID is difficult to stably control, with a long control adjustment process and high production energy consumption. Summary of the Invention

[0003] In view of the above problems, this application provides a chemical pulping cooking and washing control method and storage medium based on the ACE process, which solves the problem that the existing chemical pulping cooking process using conventional PID control can only perform single-loop control and it is difficult to achieve multi-variable fine-tuning system online optimization control.

[0004] To achieve the above object, the inventor provides a chemical pulping cooking and washing control method based on the ACE process, including:

[0005] Collect the process parameter data of the chemical pulping cooking and washing process, where the process parameter data includes cooking temperature, cooking pressure, pH value, cooking time, liquor ratio, additive concentration, steam flow rate, and liquid level;

[0006] Perform preprocessing on the collected process parameter data, and the preprocessing includes data screening and data filtering;

[0007] Identify the time delay between the process parameter data in the chemical pulping cooking and washing process;

[0008] Build a high-precision multi-variable model to capture the strong non-linearity and coupling between the process parameter data in the chemical pulping cooking and washing process;

[0009] Generate the optimal adjustment strategy of the process parameter data for the required target through the multi-variable predictive control model.

[0010] In some embodiments, the data screening includes the following steps:

[0011] The data in the steady-state interval of the process parameter data are retained, and outliers exceeding ±3σ are eliminated.

[0012] In some embodiments, the data filtering comprises the following steps:

[0013] The high-frequency noise of the process parameter data is processed by moving average filtering or bandpass filtering.

[0014] In some embodiments, the following steps are also included:

[0015] The high-precision multivariate model is simplified and verified to obtain a simplified high-precision multivariate model.

[0016] In some embodiments, the following steps are also included:

[0017] The parameters of the multivariable predictive control model are updated in real time by using an extended Kalman filter.

[0018] Another technical solution is provided, a storage medium, wherein the storage medium stores a computer program, and when the computer program is executed by a processor, the following steps are performed:

[0019] Collecting process parameter data of chemical pulping cooking and washing process, the process parameter data includes cooking temperature, cooking pressure, pH value, cooking time, liquid ratio, additive concentration, steam flow and liquid level;

[0020] Preprocessing the collected process parameter data, wherein the preprocessing includes data screening and data filtering;

[0021] Identify time delays between process parameter data in chemical pulping cooking and washing processes;

[0022] Construct a high-precision multivariate model to capture the strong nonlinearity and coupling between process parameter data of chemical pulping cooking and washing processes;

[0023] Through the multivariable predictive control model, the optimal adjustment strategy of process parameter data for the required target is generated.

[0024] In some embodiments, the data screening includes the following steps:

[0025] The data in the steady-state interval of the process parameter data are retained, and outliers exceeding ±3σ are eliminated.

[0026] In some embodiments, the data filtering comprises the following steps:

[0027] The high-frequency noise of the process parameter data is processed by moving average filtering or bandpass filtering.

[0028] In some embodiments, the following steps are further included:

[0029] Simplify and verify the high-precision multivariable model to obtain the simplified high-precision multivariable model.

[0030] In some embodiments, the following steps are further included:

[0031] Update the parameters of the multivariable predictive control model in real time by using the extended Kalman filter.

[0032] Different from the prior art, in the above technical solution, during the chemical pulping cooking and washing process, process parameter data is collected, and the collected process parameter data is preprocessed, including data screening and data filtering. Then, the time delay between the process parameter data in the chemical pulping cooking and washing process is identified, a high-precision multivariable model is constructed to capture the strong nonlinearity and coupling between the process parameter data. After that, through the multivariable predictive control module, an optimal adjustment strategy for the process parameter data of the required target is generated; by considering the nonlinearity and coupling between each process parameter data, an optimal adjustment strategy for the process parameter data is generated, avoiding the influence of single variable adjustment on other variables, and further reducing the large losses in production process in terms of output, quality, and energy consumption.

[0033] The above relevant description of the invention content is only an overview of the technical solution of this application. In order to enable those of ordinary skill in the art to more clearly understand the technical solution of this application, and then can be implemented according to the content recorded in the description and the drawings, and in order to make the above objects, other objects, features, and advantages of this application more easily understood, the following is described in conjunction with the specific embodiments and drawings of this application. Description of the Drawings

[0034] The drawings are only used to show the principles, implementation methods, applications, features, and effects of the specific embodiments of this application and other related contents, and should not be considered as a limitation to this application.

[0035] In the drawings of the specification:

[0036] Figure 1 It is a schematic structural diagram of a chemical pulping cooking and washing control method based on the ACE process described in the specific embodiment;

[0037] Figure 2 It is another schematic structural diagram of a chemical pulping cooking and washing control method based on the ACE process described in the specific embodiment;

[0038] Figure 3 It is a schematic principle diagram of a multivariable predictive control model described in the specific embodiment;

[0039] Figure 4It is a schematic structural diagram of the storage medium described in the specific implementation mode.

[0040] The descriptions of the reference numerals involved in the above-mentioned drawings are as follows:

[0041] 410. Storage medium

[0042] 420. Processor Specific implementation mode

[0043] To illustrate in detail the possible application scenarios, technical principles, specific implementable solutions, achievable purposes and effects of this application, the following will be described in detail in conjunction with the listed specific examples and with reference to the drawings. The embodiments described herein are only used to more clearly illustrate the technical solutions of this application, so they are only examples and cannot be used to limit the protection scope of this application.

[0044] Referring to "embodiments" in this text means that the specific features, structures or characteristics described in conjunction with the embodiments can be included in at least one embodiment of this application. The term "embodiment" appearing in various positions in the specification does not necessarily refer to the same embodiment, nor is it particularly limited to the independence or relevance between other embodiments. In principle, in this application, as long as there is no technical contradiction or conflict, the technical features mentioned in each embodiment can be combined in any way to form corresponding implementable technical solutions.

[0045] Unless otherwise defined, the meanings of the technical terms used in this text are the same as those generally understood by those skilled in the technical field to which this application belongs; the use of the relevant terms in this text is only for describing specific embodiments and is not intended to limit this application.

[0046] In the description of this application, the term "and / or" is an expression used to describe the logical relationship between objects, indicating that there can be three relationships. For example, A and / or B means: there is A, there is B, and there is both A and B at the same time. In addition, the character " / " in this text generally represents an "or" logical relationship between the associated objects before and after.

[0047] In this application, terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual quantity, primary or secondary, or order relationship between these entities or operations.

[0048] Without further limitations, in this application, the terms "including", "comprising", "having" or other similar expressions used in a statement are intended to cover non-exclusive inclusion. These expressions do not exclude the possibility that there may be additional elements in a process, method or product that includes the described elements. Thus, a process, method or product that includes a series of elements may include not only those defined elements, but also other elements not explicitly listed, or elements inherent to such a process, method or product.

[0049] Similar to the understanding in the "Examination Guidelines", in this application, expressions such as "greater than", "less than", "exceeding" are understood to exclude the corresponding number; expressions such as "above", "below", "within" are understood to include the corresponding number. In addition, in the description of the embodiments of this application, the meaning of "a plurality of" is two or more (including two). Similar expressions related to "many", such as "multiple groups", "multiple times", etc., are understood in the same way, unless otherwise specifically defined.

[0050] In the description of the embodiments of this application, spatial-related expressions such as "center", "longitudinal", "transverse", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "perpendicular", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", "axial", "radial", "circumferential", etc. indicate the orientation or positional relationship based on the orientation or positional relationship shown in the specific embodiment or the attached drawings. This is only for the convenience of describing the specific embodiments of this application or facilitating the understanding of the reader, rather than indicating or implying that the device or component referred to must have a specific position, a specific orientation, or be constructed or operated in a specific orientation. Therefore, it should not be construed as a limitation to the embodiments of this application.

[0051] Unless otherwise clearly specified or limited, in the description of the embodiments of this application, terms such as "installed", "connected", "joined", "fixed", "set" should be understood in a broad sense. For example, the "connection" can be a fixed connection, a detachable connection, or an integral setting; it can be a mechanical connection, an electrical connection, or a communication connection; it can be directly connected, or indirectly connected through an intermediate medium; it can be the communication inside two elements or the interaction relationship between two elements. For those skilled in the art to which this application pertains, the specific meanings of the above terms in the embodiments of this application can be understood according to specific circumstances.

[0052] Please refer to Figure 1 , a chemical pulping cooking and washing control method based on the ACE process in this embodiment includes:

[0053] Collecting process parameter data of chemical pulping cooking and washing process, the process parameter data includes cooking temperature, cooking pressure, pH value, cooking time, liquid ratio, additive concentration, steam flow and liquid level;

[0054] Preprocessing the collected process parameter data, wherein the preprocessing includes data screening and data filtering;

[0055] Identify time delays between process parameter data in chemical pulping cooking and washing processes;

[0056] Construct a high-precision multivariate model to capture the strong nonlinearity and coupling between process parameter data of chemical pulping cooking and washing processes;

[0057] Through the multivariable predictive control model, the optimal adjustment strategy of process parameter data for the required target is generated.

[0058] In the process of chemical pulping, cooking and washing, process parameter data are collected and preprocessed, including data screening and data filtering. Then, the time delay between process parameter data in the process of chemical pulping, cooking and washing is identified, and a high-precision multivariate model is constructed to capture the strong nonlinearity and coupling between process parameter data. After that, the optimal adjustment strategy of the process parameter data for the required target is generated through the multivariate predictive control module. By considering the nonlinearity and coupling between each process parameter data, the optimal adjustment strategy of the process parameter data is generated to avoid the influence of a single variable adjustment on other variables, thereby reducing the large losses in output, quality and energy consumption in the production process.

[0059] In some embodiments, the data screening includes the following steps:

[0060] Keep the data of the steady-state interval of the process parameter data.

[0061] In order to eliminate outliers in the data (such as feed blockage), the collected process parameter data are screened to retain the steady-state interval data of process parameter data such as cooking temperature, pressure, pH value, etc., for example, outliers exceeding ±3σ are eliminated.

[0062] In some embodiments, the data filtering comprises the following steps:

[0063] The high-frequency noise of the process parameter data is processed by moving average filtering or bandpass filtering.

[0064] In order to further eliminate noise interference, such as sensor offset and steam fluctuation, the collected process parameter data is filtered, and moving average filtering (MA) or bandpass filtering (BPF) is used to process high-frequency noise (such as flow sensor jitter); for example, in the cooking scenario, the pH measurement signal is stabilized by filtering to avoid control lag caused by electrode corrosion.

[0065] In some embodiments, the following steps are further included:

[0066] Simplify and verify the high-precision multivariable model to obtain a simplified high-precision multivariable model.

[0067] Simplify the high-order model to reduce the computational complexity while ensuring the control performance.

[0068] The simplification method is as follows:

[0069] Principal Component Analysis (PCA): Extract the dominance of key variables (such as temperature, pH value);

[0070] Singular Value Decomposition (SVD): Eliminate redundant state variables (such as the strong correlation between liquor ratio and cooking time).

[0071] Verification method: Verify the robustness of the reduced-order model through Monte Carlo simulation (such as adding ±10% model parameter perturbations).

[0072] In some embodiments, the following steps are further included:

[0073] Update the parameters of the multivariable predictive control model in real time by using the Extended Kalman Filter.

[0074] To achieve adaptive parameter correction, an online learning mechanism is adopted, and the Extended Kalman Filter (EKF) is used to update the model parameters (such as the reaction rate constant k) in real time. For example, when the wood batch changes, the k value in the delignification rate model is automatically corrected.

[0075] Chemical pulping cooking is a typical multivariable and strongly coupled process, involving the coordinated control of multiple parameters such as temperature, pressure, pH value, and flow rate. The ACE advanced algorithm control technology is about how to find a balance among these variables, such as how to save energy and reduce consumption while ensuring the cooking effect, or how to improve the pulp quality.

[0076] The ACE (Adaptability, Control, Energy Optimization) advanced algorithm control technology combines the process characteristics and control of the cooking and washing processes as follows:

[0077] 1. Overview of the chemical pulping cooking process

[0078] Chemical pulping is a process of decomposing lignocellulose with alkaline or acidic reagents (such as NaOH, H2O2) under high temperature and high pressure conditions. The core goal is to achieve delignification and softening of the fibers while retaining cellulose and hemicellulose as much as possible.

[0079] Key process parameters:

[0080] Temperature (90–180 °C)

[0081] Pressure (0.5–1.5 MPa)

[0082] pH value (12–14)

[0083] Cooking time (30–120 minutes)

[0084] Liquid-to-fiber ratio (mass ratio of liquid to fiber)

[0085] Additive concentration (such as sulfide, bleaching agent)

[0086] Process challenges:

[0087] Strong coupling: Temperature, pressure, and pH value affect each other (e.g., increasing temperature accelerates chemical reactions, and increasing pressure raises the boiling point).

[0088] Nonlinear dynamics: The lignin decomposition rate changes exponentially with temperature and there is a hysteresis effect.

[0089] Multi-objective optimization: It is necessary to balance pulp quality (degree of delignification, residue rate), energy consumption (steam consumption), and environmental friendliness (wastewater discharge).

[0090] Raw material fluctuations: Differences in wood species, humidity, and pretreatment lead to the need for dynamic adjustment of process parameters.

[0091] 2. The core objectives of the ACE control technology in the cooking process:

[0092] 2.1 Control accuracy

[0093] Synchronously optimize multiple variables: Maintain the stability of temperature and pressure (prevent bursting or undercooking), and precisely control the pH value to reduce chemical consumption.

[0094] Optimize the cooking curve: Automatically adjust the time allocation in the heating-up and holding stages according to the raw material characteristics.

[0095] 2.2 Adaptability

[0096] Adapt to raw materials: Cope with changes in initial conditions caused by wood species (softwood / hardwood) and batch differences.

[0097] Compensate for equipment aging: Correct sensor drift or valve hysteresis through online parameter identification.

[0098] 2.3 Improve energy efficiency

[0099] Optimize steam reuse: Dynamically adjust the steam flow rate to reduce ineffective heating (e.g., predict steam demand through MPC).

[0100] Recover heat energy: Use waste heat to preheat the feed or process water.

[0101] 2.4 Environmental Protection and Safety

[0102] Wastewater Discharge Control: Reduce chemical agent residues through closed-loop pH control.

[0103] Overpressure Protection: Prevent safety accidents caused by overpressure in the digester.

[0104] 3. Key Algorithms and Technical Implementations

[0105] 3.1 Multivariable Model Predictive Control (MPC)

[0106] Model Construction:

[0107] Mechanistic Model: Based on chemical reaction kinetics (e.g., the rate of delignification described by the Arrhenius equation)

[0108] and heat transfer equations.

[0109] Data-driven Model: Use historical data to train LSTM or CNN networks to capture non-linear coupling relationships (e.g., the lag effect of temperature on pH).

[0110] Rolling Optimization:

[0111] Take energy consumption and pulp quality as multi-objective functions, and the constraints include safety limits (e.g., pressure ≤

[0112] 1.5 MPa) and environmental protection requirements (pH ≥ 12).

[0113] Example: By optimizing the steam injection rate and liquor ratio, shorten the cooking time while reducing steam consumption.

[0114] 3.2 Adaptive Control (AC)

[0115] Parameter Self-tuning: Based on online least squares (OLS) or extended Kalman filter (EKF)

[0116] Update model parameters (such as reaction rate constants) in real time.

[0117] Model Reference Adaptive Control (MRAC): Dynamically match the reference model (ideal cooking curve) with the actual process to compensate for performance degradation caused by equipment aging.

[0118] 3.3 Decoupling Control Design

[0119] Feedforward-Feedback Decoupling:

[0120] Feedforward Compensation: Adjust the steam flow through feed flow prediction to eliminate the coupling caused by raw material changes.

[0121] Feedback Decoupling: Use a state observer to reconstruct the decoupling controller (e.g., pole-zero configuration based on transfer function matrix).

[0122] 3.4 Intelligent Optimization Strategy

[0123] Deep Reinforcement Learning (DRL): Train the policy network to directly optimize the long-term energy consumption target (such as the "cooking time - energy consumption" trade-off).

[0124] Multi-Objective Particle Swarm Optimization (MOPSO): Optimize the delignification rate, energy consumption, and wastewater discharge simultaneously to generate a Pareto optimal solution set.

[0125] 4. Control System Architecture

[0126] plaintext

[0127] Sensor Layer

[0128] ├── Temperature, pressure, pH value, and flow sensors

[0129] ├── On-line analyzer for cooking liquor (such as near-infrared spectroscopy)

[0130] └── Actuator Layer

[0131] ├── Steam control valve

[0132] ├── pH value neutralization system

[0133] └── Liquid level controller

[0134] Control Layer

[0135] ├── Data preprocessing (filtering, anomaly detection)

[0136] ├── MPC controller (rolling optimization module)

[0137] ├── Adaptive parameter estimation module

[0138] └── Safety interlock protection (such as emergency pressure relief when pressure exceeds the limit)

[0139] Optimization Layer

[0140] ├── Short-term MPC (second-level to minute-level control)

[0141] ├── Medium- and long-term optimization (cooking curve planning)

[0142] └── Energy management strategy (steam reuse scheduling)

[0143] 5. Typical Application Cases;

[0144] 5.1 Optimization of Delignification Rate

[0145] Problem: Traditional control leads to fluctuations in the delignification rate (85% - 92%), affecting the pulp quality.

[0146] - ACE Process:

[0147] Modeling of the cooking and washing process based on the LSTM network to predict the delignification rate under different temperature-time combinations.

[0148] Real-time adjustment of the steam injection strategy through MPC to stabilize the delignification rate at 91% ± 1% and reduce the waste pulp rate by 5%.

[0149] The ACE control technology for the chemical pulping cooking process significantly improves the control accuracy and energy efficiency level through multivariable collaborative optimization, adaptive learning, and energy flow reconstruction. In the future, with the integration of technologies such as digital twin and federated learning, cross-factory knowledge sharing and autonomous optimization can be achieved, promoting the development of the pulping industry towards green and low-carbon directions.

[0150] In the cooking process, the core of the multivariable and multi-collaborative control technology is to achieve multi-objective optimization (quality, energy consumption, environmental protection) through data-driven + mechanism modeling. Its architecture can be divided into three levels:

[0151] Data preprocessing layer

[0152] ├── Data screening → Data filtering → Time delay identification → High-order identification

[0153] ├── Model reduction → Model verification

[0154] └── Control optimization layer

[0155] ├── Multivariable predictive control (MPC)

[0156] ├── Adaptive parameter correction

[0157] └── Energy flow collaborative optimization

[0158] 2. The flow chart of the chemical pulping cooking and washing control method based on the ACE process is as Figure 2 shown and includes the following steps:

[0159] 2.1 Data screening and filtering

[0160] Purpose: Eliminate noise interference (such as sensor drift, steam fluctuations) and outliers (such as feed blockages).

[0161] Implementation:

[0162] Data screening: Retain the data in the steady-state intervals of cooking temperature, pressure, and pH value (such as removing the outlier points beyond ±3σ).

[0163] Data filtering: Use moving average filtering (MA) or band-pass filtering (BPF) to process high-frequency noise (such as flow sensor jitter).

[0164] Cooking scenario: For example, stabilize the pH value measurement signal through filtering to avoid control lag caused by electrode corrosion.

[0165] 2.2 Time-delay identification

[0166] Purpose: Identify the time delay between variables in the cooking and washing processes (such as the lag from steam injection to temperature rise).

[0167] Method:

[0168] Open-loop experiment: Use the step response method to measure the delay of temperature to steam flow rate (typical value: 5 - 15 seconds).

[0169] Dynamic Time Warping (DTW): Process non-linear coupling delays (such as the lag effect of lignin decomposition rate changing with temperature).

[0170] Cooking scenario: Compensate for the time difference between liquor ratio adjustment and pH value change to avoid premature or late neutralization.

[0171] 2.3 High-order model identification

[0172] Purpose: Construct a high-precision multi-variable model to capture the strong non-linearity and coupling characteristics of the cooking and washing processes.

[0173] Method:

[0174] Mechanistic model: Based on chemical reaction kinetics (such as the Arrhenius equation) + heat transfer model.

[0175] Data-driven model: Use LSTM or Transformer networks to learn the temporal correlation of temperature - pressure - pH value.

[0176] Cooking scenario: For example, establish a non-linear relationship model between temperature and delignification rate (R 2 > 0.95)

[0177] 2.4 Model reduction and verification

[0178] Purpose: Simplify the high-order model to reduce computational complexity while ensuring control performance.

[0179] - Method:

[0180] Principal Component Analysis (PCA): Extract the dominant factors of key variables (such as temperature, pH value).

[0181] Singular Value Decomposition (SVD): Eliminate redundant state variables (such as the strong correlation between liquor ratio and cooking time).

[0182] Verification: Verify the robustness of the reduced-order model through Monte Carlo simulation (such as adding ±10% perturbation to the model parameters).

[0183] 2.5 Control Optimization Layer Implementation

[0184] (1) Multivariable Predictive Control (MPC)

[0185] Rolling optimization objective:

[0186] minJ = Σ(t = 1~T) ω1ε_T 2 +ω2ΔU 2

[0187] ε_T`: Error between the delignification rate and the set value

[0188] ΔU`: Adjustment range of steam flow rate

[0189] Weights ω1, ω2`: Balance quality and energy consumption

[0190] Constraint conditions:

[0191] Safety constraint: Pressure ≤ 1.5 MPa, pH ≥ 12

[0192] Environmental protection constraint: Wastewater COD discharge ≤ 500 mg / L

[0193] (2) Adaptive Parameter Calibration

[0194] Online learning mechanism:

[0195] Use the Extended Kalman Filter (EKF) to update the model parameters (such as the reaction rate constant k) in real time.

[0196] Example: When the wood batch changes, automatically correct the k value in the delignification rate model.

[0197] (3) Energy Flow Collaborative Optimization

[0198] Multi-objective optimization strategy:

[0199] Objective function: min(steam consumption, wastewater discharge, control delay)

[0200] Decision variables: Steam flow rate, liquor ratio, neutralizer injection volume

[0201] Solution tool: The NSGA-II algorithm generates the Pareto optimal solution set.

[0202] 3. Typical Application Cases;

[0203] Case 1: Dynamic Optimization of Cooking Curve

[0204] Problem: The traditional fixed cooking program cannot adapt to different wood species.

[0205] ACE Solution:

[0206] 1. Obtain high-quality process data through data screening and filtering.

[0207] 2. Use the LSTM network to establish a mapping model from raw materials (coniferous wood / hardwood) to cooking curves.

[0208] 3. Dynamically generate optimal heating and insulation time series based on MPC.

[0209] Effect: The standard deviation of delignification rate was reduced from 8% to 3%, and energy consumption was reduced by 10%.

[0210] Case 2: Pressure-Temperature Decoupling Control

[0211] Problem: Steam pressure fluctuations causing unstable temperature control.

[0212] ACE Solution:

[0213] 1. Separate the coupling terms of pressure and temperature through time delay identification.

[0214] 2. Design a feedforward-feedback decoupling controller:

[0215] Feed Forward: Predicts steam demand based on feed flow.

[0216] Feedback: Tune PID parameters based on reduced-order model.

[0217] Effect: Temperature control deviation is reduced from ±2℃ to ±0.5℃, and steam waste is reduced by 15%.

[0218] Through the data processing and model reduction steps shown in the flowchart, ACE multi-variable multi-cooperative control technology has achieved the following breakthroughs:

[0219] 1. Data-driven + mechanism fusion: balance model accuracy and computational efficiency.

[0220] 2. Dynamic adaptation: respond to changes in raw materials and equipment in real time.

[0221] 3. Multi-objective collaborative optimization: taking into account the needs of process quality, output, energy consumption and environmental protection.

[0222] Case 3: Wood chip bin level control technology:

[0223] Important note: Stable chip bin level control allows for better pre-steaming of chips, more air removal, and more effective white liquor penetration. This will facilitate chip level control in the digester, thereby improving the pulp quality of the digester.

[0224] Control scheme: Control the chip bin level at a desired set point SP, thereby linking with precise control of the chip metering screw speed to predict production changes caused by changes in chip consumption or by cooking chip level control.

[0225] Case 4: Chip level control technology for the cooking tower:

[0226] Important note: Reasonably controlling the chip level in the cooking tower can improve the movement of chips in the tower and stabilize the compaction degree of the cooking tower, improve pulp quality, and reduce the linear change of the kappa number.

[0227] Control scheme: By adjusting the speed of the chip feed metering screw and the steam flow rate, the chip level in the cooking tower is controlled at the required setpoint SP, achieving CAS cascade control, so as to make the final PV measurement value consistent with the initial SP setpoint and make the entire system PID stable.

[0228] Case 5: Slurry level control technology for the cooking tower

[0229] Important note: Level control can reduce level fluctuations, stabilize the compaction degree of the cooking tower column, improve the operability of cooking tower level control, and thus reduce the fluctuating change of the kappa number.

[0230] Control scheme: By adjusting the upper extraction (also known as low solids extraction) flow rate, the cooking tower level is controlled at the desired setpoint, while keeping the upper extraction differential pressure below the desired maximum value.

[0231] Case 6: Upper / lower extraction residual alkali control

[0232] Important note: (Upper / lower) Residual alkali control calculation corrects the ratio target of alkali / chips; by more precise residual alkali control, the consumption of white liquor can be saved.

[0233] Control scheme: (Upper / lower) This control strategy maintains the residual alkali amount of the upper / lower extraction (also known as low solids extraction) by sending a correction instruction to the AW ratio control, and then adjusts the flow of white liquor to the chip cooking circulation pipeline.

[0234] Case 7: Kappa number and H-factor control

[0235] Important note: The kappa number is the most important quality parameter in the cooking and washing process. Controlling it determines the required temperature target value for the cooking tower, thus ensuring that the H-factor remains at the target value. Precise temperature control can compensate for the change in cooking residence time, thus reducing the change in the kappa number.

[0236] Control scheme: By adjusting the H-factor in the cooking zone, the kappa number at the spraying line is controlled at the required SP setpoint. The H-factor is controlled by adjusting the temperature of the cooking circulation heater.

[0237] Case 8: Downflow and dilution factor control

[0238] Important Note: The dilution factor of the cooking tower represents the ratio between the cold blowdown flow rate (i.e., the dilution flow rate applied to the bottom of the digester) and the production rate of the cooking tower. If the dilution factor of the cooking tower is not properly controlled, changes in column movement, the chip level in the cooking tower, the Kappa number, and the blowdown concentration may occur. Most importantly, a stable dilution factor can improve the washing effect in the digester, thereby reducing the water consumption in the washing area, higher evaporation of black liquor solids, and more stable operation of the entire wood fiber line. The dilution factor is controlled at an ideal target by adjusting the total blowdown flow rate. The total cold blowdown consists of the vertical dilution flow rate and the horizontal dilution flow rate. The vertical dilution flow rate is used to control the blowdown pressure difference. Therefore, the horizontal dilution flow rate is determined by the difference between the total blowdown flow rate and the vertical dilution flow rate. Downflow is a term used to indicate that there is a downward flowing free liquid (solution other than chips) in the washing area on the lower extraction screen. This downward free liquid is important for improving the movement of chips in the tower, enhancing the operability of the cooking tower, and reducing changes in the Kappa number.

[0239] Control Scheme: The downflow factor is controlled at an ideal target by adjusting the lower extraction flow rate. If the lower extraction pressure difference increases, the control device can adjust the dilution factor.

[0240] Among them, the principle of the multivariable predictive control model in the chemical pulping cooking and washing control method based on the ACE process is as Figure 3 shown, including the following three steps: (1) Predict the future dynamics of the system; (2) Numerically solve the open-loop optimization problem; (3) Apply the first element of the optimized solution to the system.

[0241] Advantage 1: Achieve precise large closed-loop PID control

[0242] Advantage 2: Actual precise automatic control, reducing the labor intensity of employees' operations

[0243] Advantage 3: Improve production output and quality, reduce steam consumption, and play a role in energy conservation and consumption reduction

[0244] Please refer to Figure 4 , a storage medium 410, the storage medium 410 stores a computer program, and when the computer program is run by a processor 420, it executes the following steps:

[0245] Collect process parameter data of the chemical pulping cooking and washing process, and the process parameter data includes cooking temperature, cooking pressure, pH value, cooking time, liquor ratio, additive concentration, steam flow rate, and liquid level;

[0246] Preprocess the collected process parameter data, and the preprocessing includes data screening and data filtering;

[0247] Identify the time delay between process parameter data in the chemical pulping cooking and washing process;

[0248] Construct a high-precision multivariate model to capture the strong nonlinearity and coupling between process parameter data of chemical pulping cooking and washing processes;

[0249] Through the multivariable predictive control model, the optimal adjustment strategy of process parameter data for the required target is generated.

[0250] In the process of chemical pulping, cooking and washing, process parameter data are collected and preprocessed, including data screening and data filtering. Then, the time delay between process parameter data in the process of chemical pulping, cooking and washing is identified, and a high-precision multivariate model is constructed to capture the strong nonlinearity and coupling between process parameter data. After that, the optimal adjustment strategy of the process parameter data for the required target is generated through the multivariate predictive control module. By considering the nonlinearity and coupling between each process parameter data, the optimal adjustment strategy of the process parameter data is generated to avoid the influence of a single variable adjustment on other variables, thereby reducing the large losses in output, quality and energy consumption in the production process.

[0251] In some embodiments, the data screening includes the following steps:

[0252] Keep the data of the steady-state interval of the process parameter data.

[0253] In order to eliminate outliers in the data (such as feed blockage), the collected process parameter data are screened to retain the steady-state interval data of process parameter data such as cooking temperature, pressure, pH value, etc., for example, outliers exceeding ±3σ are eliminated.

[0254] In some embodiments, the data filtering comprises the following steps:

[0255] The high-frequency noise of the process parameter data is processed by moving average filtering or bandpass filtering.

[0256] In order to further eliminate noise interference, such as sensor offset and steam fluctuation, the collected process parameter data is filtered, and moving average filtering (MA) or bandpass filtering (BPF) is used to process high-frequency noise (such as flow sensor jitter); for example, in the cooking scenario, the pH measurement signal is stabilized by filtering to avoid control lag caused by electrode corrosion.

[0257] In some embodiments, the following steps are also included:

[0258] The high-precision multivariate model is simplified and verified to obtain a simplified high-precision multivariate model.

[0259] Simplify high-order models to reduce computational complexity while maintaining control performance.

[0260] The simplified method is as follows:

[0261] Principal Component Analysis (PCA): Extract the dominance of key variables (such as temperature, pH value);

[0262] Singular Value Decomposition (SVD): Eliminate redundant state variables (such as the strong correlation between liquor ratio and cooking time).

[0263] Verification method: Verify the robustness of the reduced-order model through Monte Carlo simulation (such as adding ±10% perturbation to the model parameters).

[0264] In some embodiments, the following steps are further included:

[0265] Update the parameters of the multivariable predictive control model in real time by using the Extended Kalman Filter.

[0266] To achieve adaptive parameter correction, an online learning mechanism is adopted, and the Extended Kalman Filter (EKF) is used to update the model parameters (such as the reaction rate constant k) in real time. For example, when the wood batch changes, the k value in the delignification rate model is automatically corrected.

[0267] Finally, it should be noted that although the above embodiments have been described in the text and drawings of the specification of this application, the patent protection scope of this application cannot be limited thereby. Any technical solutions obtained by equivalent structure or equivalent process substitution or modification based on the essential concept of this application, using the content recorded in the text and drawings of the specification of this application, and any technical solutions directly or indirectly implementing the above embodiments in other related technical fields, etc., are all included in the patent protection scope of this application.

Claims

1. A chemical pulping cooking and washing control method based on the ACE process, characterized in that, Including: Collecting process parameter data of the chemical pulping cooking and washing process, where the process parameter data includes cooking temperature, cooking pressure, pH value, cooking time, liquor ratio, additive concentration, steam flow rate, and liquid level; Preprocessing the collected process parameter data, where the preprocessing includes data screening and data filtering; Identifying the time delay between process parameter data in the chemical pulping cooking and washing process; Constructing a high-precision multivariable model to capture the strong nonlinearity and coupling between process parameter data in the chemical pulping cooking and washing process; Generating an optimal adjustment strategy for process parameter data of the required target through a multivariable predictive control model.

2. The chemical pulping cooking and washing control method based on the ACE process according to claim 1, wherein The data screening includes the following steps: Retaining the data in the steady-state interval of the process parameter data.

3. The chemical pulping cooking and washing control method based on the ACE process according to claim 1, wherein, The data filtering includes the following steps: Processing the high-frequency noise of the process parameter data through moving average filtering or band-pass filtering.

4. The chemical pulping cooking and washing control method based on the ACE process according to claim 1, wherein It also includes the following steps: Simplifying and validating the high-precision multivariable model to obtain a simplified high-precision multivariable model.

5. The chemical pulping cooking and washing control method based on the ACE process according to claim 1, wherein It also includes the following steps: Real-time updating the parameters of the multivariable predictive control model by using the extended Kalman filter.

6. A storage medium storing a computer program, characterized in that, When the computer program is run by a processor, it performs the following steps: Collecting process parameter data of the chemical pulping cooking and washing process, where the process parameter data includes cooking temperature, cooking pressure, pH value, cooking time, liquor ratio, additive concentration, steam flow rate, and liquid level; Preprocessing the collected process parameter data, where the preprocessing includes data screening and data filtering; Identifying the time delay between process parameter data in the chemical pulping cooking and washing process; Constructing a high-precision multivariable model to capture the strong nonlinearity and coupling between process parameter data in the chemical pulping cooking and washing process; Generating an optimal adjustment strategy for process parameter data of the required target through a multivariable predictive control model.

7. The storage medium according to claim 6, characterized in that, The data screening includes the following steps: Retaining the data in the steady-state interval of the process parameter data.

8. The storage medium according to claim 6, wherein The data filtering includes the following steps: Processing the high-frequency noise of the process parameter data through moving average filtering or band-pass filtering.

9. The storage medium according to claim 6, wherein It also includes the following steps: Simplifying and validating the high-precision multivariable model to obtain a simplified high-precision multivariable model.

10. The storage medium according to claim 6, characterized in that, It also includes the following steps: Real-time updating the parameters of the multivariable predictive control model by using the extended Kalman filter.

Images