Process dynamic optimization method and system based on carbon capture platform

Abstract

The application discloses a process dynamic optimization method and system based on a carbon capture platform, and relates to the technical field of carbon capture. The method comprises the following steps: acquiring a demand output and a carbon capture process parameter space for current cement production; adjusting a first carbon capture process parameter in the carbon capture process parameter space, combining the demand output, and predicting a first predicted carbon capture parameter and a first calciner temperature; performing cement production influence prediction according to the first calciner temperature and the demand output, and obtaining a first predicted influence output; calculating a first carbon capture score according to the first predicted carbon capture parameter and the first predicted influence product, adjusting a first adjustment step according to the first predicted influence output, obtaining a second adjustment step, and continuing to iteratively optimize the carbon capture process parameter to obtain optimal carbon capture process parameters, and the full-oxygen combustion unit and the calcium cycle capture unit are controlled. The application effectively improves the stability of cement production capacity.

Description

Technical Field

[0001] This invention relates to the field of carbon capture technology, and more specifically to a method and system for dynamic process optimization based on a carbon capture platform. Background Technology

[0002] With the rapid advancement of global industrialization, carbon dioxide emissions have become a focus of international attention. The cement industry, as a typical high-energy-consuming and high-emission sector, accounts for a significant proportion of global carbon dioxide emissions, necessitating emission reduction through carbon capture technologies. Currently common carbon capture technologies mainly include post-combustion capture, oxy-fuel combustion, and calcium cycle technology.

[0003] There is a mutually restrictive relationship between stable cement production and carbon dioxide capture efficiency: cement clinker calcination requires a stable high temperature of around 1450℃. This temperature range ensures the efficiency and consistency of limestone decomposition and clinker mineral formation, which is the foundation for stable production. If the temperature is adjusted to improve carbon capture efficiency, the carbon capture calciner will compete with the clinker production furnace for temperature, thereby disrupting the temperature balance of clinker calcination. This leads to problems such as incomplete calcination, kiln ring formation, and material blockage, directly causing production interruptions or a decrease in clinker output per unit time, thus undermining production stability. Traditional single capture methods are difficult to balance cement production stability and carbon capture efficiency. The carbon capture effect is generally poor, and the capture efficiency is unstable, which may lead to insufficient energy efficiency and affect cement production. Summary of the Invention

[0004] This application provides a method and system for dynamic process optimization based on a carbon capture platform, aiming to solve the technical problem of poor carbon capture control effect in the prior art.

[0005] In view of the above problems, this application provides a method and system for dynamic process optimization based on a carbon capture platform.

[0006] Firstly, this application provides a method for dynamic process optimization based on a carbon capture platform, including:

[0007] Obtain the current cement production demand and the carbon capture process parameter space, where each carbon capture process parameter includes the all-oxygen combustion parameter and the calcium cycle parameter;

[0008] Based on the required output, configure the first adjustment step size, adjust and generate the first carbon capture process parameters within the carbon capture process parameter space, and combine the required output to predict and output the first predicted carbon capture parameters and the first calciner temperature.

[0009] Based on the temperature and required output of the first calcining furnace, the impact on cement production is predicted, and the first predicted impact output is obtained.

[0010] A first carbon capture score is calculated based on the first predicted carbon capture parameters and the first predicted impact on the product. Based on the first predicted impact on the output, the first adjustment step size is adjusted to obtain a second adjustment step size. The carbon capture process parameters are then iteratively optimized to obtain the optimal carbon capture process parameters, and the oxygen-fuel combustion unit and the calcium cycle capture unit are controlled.

[0011] Secondly, this application provides a process dynamic optimization system based on a carbon capture platform, including:

[0012] The parameter acquisition module is used to obtain the required output for current cement production and to obtain the carbon capture process parameter space, wherein each carbon capture process parameter includes the all-oxygen combustion parameter and the calcium cycle parameter.

[0013] The parameter prediction module is used to configure a first adjustment step size according to the demand output, adjust and generate the first carbon capture process parameters within the carbon capture process parameter space, and predict and output the first predicted carbon capture parameters and the first calciner temperature in combination with the demand output.

[0014] The production impact prediction module is used to predict the impact on cement production based on the temperature of the first calcining furnace and the required output, and to obtain the first predicted impact output.

[0015] The dynamic optimization control module is used to calculate a first carbon capture score based on the first predicted carbon capture parameters and the first predicted impact on the product, and adjust the first adjustment step size based on the first predicted impact on the output to obtain a second adjustment step size, and continue to iteratively optimize the carbon capture process parameters to obtain the optimal carbon capture process parameters, and control the all-oxygen combustion unit and the calcium cycle capture unit.

[0016] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0017] This application provides a method and system for dynamic optimization of carbon capture processes based on a carbon capture platform. It dynamically adjusts process parameters and step sizes according to real-time production demand, closely linking parameter optimization with production needs and achieving dynamic adaptation of carbon capture process parameters. Through an iterative mechanism that predicts the impact of carbon capture parameters on production and dynamically adjusts the step size, it achieves adaptive convergence of parameter optimization, balancing carbon capture efficiency and cement production to ensure stable output. The joint optimization of oxy-fuel combustion and calcium cycling improves carbon dioxide capture efficiency. Iterative calculations obtain optimal parameters, enabling intelligent control of the carbon capture process, maximizing carbon capture efficiency, ensuring cement production needs are met, and improving the overall operational performance of the carbon capture platform. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying 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.

[0019] Figure 1 A schematic flowchart of the process dynamic optimization method based on a carbon capture platform provided in the embodiments of this application;

[0020] Figure 2 A schematic diagram of the structure of the process dynamic optimization system based on a carbon capture platform provided in the embodiments of this application;

[0021] The components represented by each number in the attached diagram are explained below:

[0022] Parameter acquisition module 11, parameter prediction module 12, production impact prediction module 13, dynamic optimization control module 14. Detailed Implementation

[0023] This application provides a process dynamic optimization method and system based on a carbon capture platform, which is used to address the technical problem of poor carbon capture control in the prior art.

[0024] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0025] It should be noted that the terms "comprising" and "having" are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to these processes, methods, products, or devices.

[0026] Example 1, as Figure 1 As shown, this application provides a process dynamic optimization method based on a carbon capture platform, the method comprising:

[0027] S100: Obtain the current cement production demand and the carbon capture process parameter space, wherein each carbon capture process parameter includes the all-oxygen combustion parameter and the calcium cycle parameter.

[0028] In this embodiment, the required output for current cement production is obtained, along with a carbon capture process parameter space. Each carbon capture process parameter includes oxy-fuel combustion parameters and calcium cycle parameters. Obtaining the current required cement production output ensures optimal carbon capture efficiency while meeting the predetermined production plan. The process parameters of the oxy-fuel combustion and calcium cycle units cannot be adjusted indefinitely. A predefined process parameter space, i.e., the physically permissible adjustment range of each parameter, ensures that all subsequent parameter combinations are engineering-feasible and within safe operating limits.

[0029] Step S100 in the method provided in this application embodiment includes:

[0030] Obtain the current required output for cement production;

[0031] The oxygen flow rate adjustment space for all-oxygen combustion and the adsorbent circulation rate space for calcium-cycled carbon capture are obtained. These are combined to obtain the carbon capture process parameter space, where each carbon capture process parameter includes all-oxygen combustion parameters and calcium circulation parameters.

[0032] In this embodiment, the current cement production demand is first obtained. Assume a cement plant's current cement production demand is 100 tons / hour. This output is used as the target constraint for optimization, limiting the adjustable range of various process parameters during carbon capture, and serving as the benchmark for all subsequent optimizations.

[0033] Secondly, the oxygen flow rate adjustment space for oxy-fuel combustion and the adsorbent circulation rate space for calcium-cycled carbon capture are obtained, and combined to obtain the carbon capture process parameter space. Each carbon capture process parameter includes oxy-fuel combustion parameters and calcium circulation parameters. For example, based on burner design and safety standards, the oxygen flow rate can be adjusted between 8000 and 12000 Nm³ / h. Below 8000 Nm³ / h, incomplete combustion may occur, while above 12000 Nm³ / h, there are safety risks. Based on reactor delivery capacity and chemical reaction kinetics, the adsorbent circulation rate can be adjusted between 50 and 90 t / h. Too low a rate results in insufficient capture, while too high a rate leads to incomplete reaction and a surge in energy consumption. Thus, a two-dimensional carbon capture process parameter space is obtained with oxygen flow rate and adsorbent circulation rate as coordinate axes: {Oxygen flow rate: 8000-12000 Nm³ / h; Adsorbent rate: 50-90 t / h}. Any valid combination of process parameters must fall within this space. This range reflects the dynamically adjustable oxygen supply under the premise of ensuring stable combustion; and reflects the allowable circulation rate range under the condition of ensuring effective calcium carbonate reaction and maintaining adsorbent activity.

[0034] In this embodiment, a process parameter space covering both oxy-fuel combustion and calcium cycle parameters is established while obtaining the required output. This parameter space is directly related to the production demand, avoiding invalid parameter searches. It takes into account both oxy-fuel combustion and calcium cycle, improving the comprehensiveness of process optimization. It provides a clear range of feasible parameters, giving subsequent prediction and iterative optimization boundary constraints and ensuring the feasibility of the results.

[0035] S200: Configure a first adjustment step size according to the required output, adjust and generate a first carbon capture process parameter within the carbon capture process parameter space, and combine the required output to predict and output a first predicted carbon capture parameter and a first calciner temperature.

[0036] In this embodiment, a first adjustment step size is configured based on the required output. First carbon capture process parameters are generated within the carbon capture process parameter space. Combined with the required output, first predicted carbon capture parameters and a first calciner temperature are predicted and output. The deviation between the required output and the historical average output directly reflects production load fluctuations. If the output is significantly higher than the average, a larger step size is needed to quickly narrow the parameter search range; if the output is close to the average, a smaller step size is needed for fine adjustment. A fixed step size cannot adapt to this fluctuation, therefore, the step size needs to be dynamically optimized through a product adjustment coefficient. Adjusting directly based on the currently operating carbon capture process parameters avoids drastic fluctuations in equipment operating conditions caused by parameter mutations, ensuring system stability. By outputting the carbon capture efficiency and calciner temperature in advance through the carbon capture predictor, the feasibility of the parameters can be predicted before they are actually put into production, avoiding output losses or emission reduction failures caused by invalid parameters.

[0037] Step S200 in the method provided in this application embodiment includes:

[0038] Get the preset adjustment step size;

[0039] The ratio of the demand output to the average demand output over a historical period is calculated and used as the product adjustment coefficient. The preset adjustment step size is then adjusted to obtain the first adjustment step size.

[0040] Obtain the current carbon capture process parameters, adjust them using the first adjustment step size, and obtain the first carbon capture process parameters;

[0041] The first carbon capture process parameters and required output are input into the carbon capture predictor, and the first predicted carbon capture parameters and the first calciner temperature are output.

[0042] In this embodiment, the preset adjustment step size is first obtained. The preset step size needs to take into account the equipment's adjustment sensitivity and the allowable fluctuation range of the process to avoid instability caused by an excessively large step size, or an excessively small step size that prolongs the optimization cycle. For example, the preset adjustment step size is: oxygen flow rate step size 400 Nm³ / h for the all-oxygen combustion unit; and adsorbent circulation rate step size 4 t / h for the calcium circulation unit.

[0043] Then, the ratio of the required output to the average required output over a historical period is calculated as the product adjustment coefficient. This coefficient is used to adjust the preset adjustment step size to obtain the first adjustment step size. For example, the historical required output of cement production for the cement plant over the past week is retrieved. The calculated average required output over the historical period is 80 t / h, and the current required output is 100 t / d. The product adjustment coefficient = current required output / historical average required output = 100 / 80 = 1.25. A coefficient > 1 indicates that the current output is higher than the normal load, requiring an increase in the step size to cope with high-load conditions and improve optimization efficiency. The first adjustment step size = preset step size × product adjustment coefficient. The first adjustment step size for oxygen flow rate in oxy-fuel combustion is 400 × 1.25 = 500 Nm³ / h; the first adjustment step size for calcium circulating adsorbent rate is 4 × 1.25 = 5 t / h.

[0044] Secondly, the current carbon capture process parameters are obtained and adjusted using the first adjustment step size to obtain the first carbon capture process parameters. For example, the current carbon capture process parameters are: oxygen flow rate 10000 Nm³ / h, adsorbent circulation rate 70 t / h, and historical average output 80 t / d. Due to the current increased demand, the oxygen flow rate and adsorbent rate need to be appropriately increased. The first all-oxygen combustion parameter = current oxygen flow rate - first adjustment step size = 10000 + 50 = 10500 Nm³ / h; the first calcium circulation parameter = current adsorbent rate - first adjustment step size = 70 + 5 = 75 t / h; the final first carbon capture process parameters are: {oxygen flow rate 10500 Nm³ / h, adsorbent circulation rate 75 t / h}.

[0045] Finally, the first carbon capture process parameters and required output are input into the carbon capture predictor, and the first predicted carbon capture parameters and the first calciner temperature are output. A neural network model is trained using historical sample data to construct the carbon capture predictor, with the process parameters and required output as inputs, and the carbon capture parameters and calciner temperature as outputs. For example, inputting the first carbon capture process parameters {oxygen flow rate 10500 Nm³ / h, adsorbent circulation rate 75 t / h} and the current required output of 4600 t / d into the model yields the following prediction results: First predicted carbon capture parameters: carbon dioxide capture efficiency 87%; First calciner temperature: 1245℃.

[0046] The steps for building the carbon capture predictor include:

[0047] First, based on historical operational data of carbon capture in cement production, sample sets of carbon capture process parameters and sample sets of required output are collected. Carbon capture rates and calciner temperatures under different sample carbon capture process parameters and required output are also collected, and these are labeled to obtain sample sets of predicted carbon capture parameters and sample sets of calciner temperatures. Historical operational data from the past year is extracted from the plant's database. Input data: Sample set of carbon capture process parameters, such as hourly recorded oxygen flow rates and adsorbent circulation rates; sample set of required output, such as planned or actual hourly cement clinker production. Output data: Sample set of predicted carbon capture parameters, such as the actual carbon capture rate measured and calculated simultaneously with the process parameters using a flue gas analyzer; sample set of calciner temperatures, such as the actual calciner temperature measured simultaneously with the process parameters using thermocouples. For example, given the inputs: {Oxygen flow rate = 10500 Nm³ / h, Adsorbent circulation rate = 75 t / h, Required output = 100 t / h}, the outputs are: {Carbon capture rate = 87%, Calciner temperature = 1245°C}. Thousands of paired data points are then compiled to form the training and test sets.

[0048] Secondly, a capture predictor is constructed based on machine learning. This capture predictor includes a capture parameter prediction branch and a calciner temperature prediction branch. A neural network structure with a shared input layer and a dual-branch output is adopted. The shared input layer receives 3-dimensional input {oxygen flow rate, adsorbent rate, and required yield} and extracts common features through two fully connected layers. The capture parameter prediction branch, after obtaining features from the shared layer, outputs the carbon capture rate through one fully connected layer and an output layer. The calciner temperature prediction branch, after obtaining features from the shared layer, outputs the temperature value through one fully connected layer and an output layer.

[0049] Finally, using the sample carbon capture process parameter set and sample demand output set as input training data, and the sample predicted carbon capture parameter set and sample calcination furnace temperature set as output supervision data, the capture predictor is trained and optimized until the test converges, completing the setup. Input training data: {sample process parameter set, sample demand output set}; Output supervision data: {sample carbon capture rate set, sample furnace temperature set}; Through iterative training and parameter optimization, the prediction error is continuously reduced; when the validation set and test set results converge, the capture predictor setup is complete. For example, after training, when a new data combination {11000, 80, 105} is input, the trained model will quickly output two predicted values, such as {88.5%, 1260°C}.

[0050] In this embodiment, the product adjustment coefficient dynamically adapts to production fluctuations, avoiding the problems of slow optimization under high load and poor accuracy under low load that occur with fixed step sizes, making the initial parameter adjustment more targeted. The first parameter is generated by fine-tuning based on the current operating parameters, avoiding the risks of sudden temperature rises and falls in the calciner and overload of the adsorbent delivery system caused by parameter mutations, thus improving the operational stability of the carbon capture platform. The carbon capture predictor verifies in advance whether the carbon capture efficiency meets the standard and whether the temperature is compliant, and selects effective initial parameters. Invalid parameters can be eliminated without actual production, reducing production loss and energy waste.

[0051] S300: Based on the temperature of the first calcining furnace and the required output, perform a prediction of the impact on cement production to obtain the first predicted impact output.

[0052] In this embodiment, the impact on cement production is predicted based on the temperature and required output of the first calciner, resulting in a first predicted impact on output. Fluctuations in calciner temperature directly affect the output and quality of clinker. By constructing a predictor through machine learning, the nonlinear relationship between temperature, required output, and output loss in historical data is learned, enabling the output of accurate output loss values. This provides a quantitative benchmark for balancing carbon capture efficiency and output, and can predict in advance whether a certain carbon capture scheme will cause a significant decrease in output due to unsatisfactory calciner temperature.

[0053] Step S300 in the method provided in this application embodiment includes:

[0054] Based on historical cement production data, we collected sample calciner temperature sets and sample demand output sets, and collected the output decreases when different sample calciner temperatures and sample demand outputs occurred, and labeled them to obtain the sample predicted output impact set.

[0055] A cement production impact predictor was constructed based on machine learning.

[0056] The cement production impact predictor is trained and optimized by using the sample calciner temperature set and the sample demand output set as input training data and the sample predicted impact output set as output supervision data until the test converges, thus completing the setup.

[0057] The temperature of the first calcining furnace and the required output are input into the cement production impact predictor, which then outputs the first predicted impact output.

[0058] In this embodiment, firstly, based on historical cement production data, sample calciner temperature sets and sample demand output sets are collected. The output decreases when different sample calciner temperatures and sample demand outputs occur are also collected, and these are labeled to obtain a sample predicted output impact set. Production records from the past two years are exported. For each time point, the impact output is calculated as: actual output output - planned demand output. The impact output is typically zero or negative, indicating output loss. All data points with non-zero impact output are selected. The sample calciner temperature set and sample demand output set are used as input features x, and the sample predicted output impact set is used as the prediction target y, forming a training dataset.

[0059] Secondly, a cement production impact predictor is constructed based on machine learning. A gradient boosting regression tree model is used to learn the mapping relationship between [calciner temperature, demand output] and [impact output]. The gradient boosting regression tree model performs excellently on tabular data and can effectively capture the complex interactions between features. The model input layer has two nodes: calciner temperature and demand output. The model consists of hundreds of decision trees, and the predictions of all decision trees are summed to output a continuous numerical value, i.e., the predicted impact output. For example, if the input temperature is <1248 and the demand output is >95, the output predicted impact is -1.5 tons.

[0060] Furthermore, the sample calcining furnace temperature set and sample demand output set are used as input training data, and the sample predicted output impact set is used as output supervision data to train and optimize the cement production impact predictor until the test converges, completing the setup. The prepared data is used to train the model and optimize its performance. For example, all collected samples are randomly shuffled and divided into training and test sets in an 80:20 ratio. The training set data is input into the model. Through continuous iteration, when the model's loss on the test set no longer decreases for 10 consecutive training epochs, or even begins to increase, training is stopped; at this point, the model is considered to have converged to its optimal generalization ability.

[0061] Finally, the temperature of the first calciner and the required output are input into the cement production impact predictor, which then outputs a first predicted impact output. For example, inputting the first calciner temperature of 1245°C and the required output of 100 t / h into the trained cement production impact predictor results in an output of: First predicted impact output = -1.5 t / h. This means that if the carbon capture process parameters {10500, 75} are implemented, although the carbon capture rate may increase to 87%, the calciner temperature will drop from the optimal 1250°C to 1245°C, and the actual cement clinker output is predicted to decrease by 1.5 t / h, i.e., the actual output is approximately 98.5 t / h, which cannot fully meet the 100 t / h requirement.

[0062] In this embodiment of the application, by constructing a cement production impact predictor, the production decline under different furnace temperature and output conditions can be quantitatively predicted: production is taken into account when optimizing carbon capture, so as to avoid serious losses caused by furnace temperature fluctuations; accurate production-side constraint indicators are provided for subsequent evaluation and iterative optimization; and the balance between carbon capture efficiency and cement production is optimized to avoid the negative impact caused by single-objective optimization.

[0063] S400: Calculate the first carbon capture score based on the first predicted carbon capture parameters and the first predicted impact on the product, and adjust the first adjustment step size based on the first predicted impact on the output to obtain the second adjustment step size. Continue to iteratively optimize the carbon capture process parameters to obtain the optimal carbon capture process parameters, and control the all-oxygen combustion unit and the calcium cycle capture unit.

[0064] In this embodiment, a first carbon capture score is calculated based on the first predicted carbon capture parameters and the first predicted impact on the product. Based on the first predicted impact on production, a first adjustment step size is adjusted to obtain a second adjustment step size. The carbon capture process parameters are then iteratively optimized to obtain the optimal carbon capture process parameters, and the all-oxygen combustion unit and calcium cycle capture unit are controlled. Carbon capture rate and production achievement rate are objectives with different or even conflicting dimensions. A comprehensive scoring function unifies these two objectives into a single scalar value, providing a clear and quantifiable basis. The next search step size is dynamically adjusted. If the impact is large, it indicates that the parameter adjustment is too aggressive, and the step size needs to be reduced for a more refined search. If the impact is small, it indicates that the current area is relatively safe, and the step size can be maintained or appropriately increased to accelerate the exploration. This step links the exploration and evaluation results, makes a decision, and adjusts the step size accordingly. Then, a new round of exploration-evaluation-decision cycle is initiated and continues until the process parameters with the best overall performance under given conditions are found.

[0065] Step S400 in the method provided in this application embodiment includes:

[0066] Based on the first prediction of the impact on product and demand output, the first actual output is calculated, and the first output achievement rate is calculated.

[0067] The first carbon capture score is calculated based on the first predicted carbon capture parameters and the first production achievement rate.

[0068] In this embodiment, firstly, based on the first predicted impact product and demand output, the first actual output is calculated, and the first output achievement rate is calculated. Continuing the previous example, the demand output is 100 t / h, the first predicted impact output is -1.5 t / h, and the first predicted carbon capture parameter is a carbon capture rate of 87%. The first actual output = demand output + first predicted impact output = 100 + (-1.5) = 98.5 t / h, and the first output achievement rate = first actual output / demand output = 98.5 / 100 = 0.985, that is, the first output achievement rate is 98.5%.

[0069] Secondly, based on the first predicted carbon capture parameters and the first production achievement rate, the first carbon capture score is calculated. The first carbon capture score = carbon capture rate × weight 1 + production achievement rate × weight 2. For example, considering the need for prioritizing emission reduction while also taking production into account, weight 1 = 0.6, weight 2 = 0.4, and the score range is 0-100 points; the first carbon capture score = 0.87 × 0.60 + 0.985 × 0.3 = 0.522 + 0.2955 = 0.8175. The first carbon capture score is a comprehensive evaluation of the first set of parameters {10500, 75} and will become the benchmark for subsequent iterative optimization. The optimization goal is to find the parameter combination that maximizes this score.

[0070] Specifically, based on the first predicted impact on production, the first adjustment step size is adjusted to obtain a second adjustment step size. The carbon capture process parameters are then iteratively optimized to obtain the optimal carbon capture process parameters, including:

[0071] First, based on the first predicted impact on output and demand output, a step size adjustment coefficient is calculated to adjust the first adjustment step size, resulting in a second adjustment step size. The step size adjustment coefficient = 1 - |first predicted impact on output / demand output|, and the second adjustment step size = first adjustment step size × step size adjustment coefficient. For example, the step size adjustment coefficient == 1 - |-1.5 / 100| = 1 - 0.015 = 0.985, the first adjustment step size is {500 Nm³ / h, 5 t / h}, and the second adjustment step size = {500 × 0.985 ≈ 492.5 Nm³ / h, 5 × 0.985 ≈ 4.93 t / h}. Since the first set of parameters caused a 98.5% output loss, more caution is needed. Therefore, the next search step size is slightly reduced to 98.5% of the original to improve optimization accuracy.

[0072] Next, the carbon capture process parameters are adjusted using the second adjustment step size to obtain the second carbon capture process parameters. Starting with the first set of parameters {10500, 75}, the carbon capture process parameters are adjusted using a reduced second adjustment step size {492.5, 4.93}. For example, reducing the oxygen flow rate generates the second carbon capture process parameters: {10500-492.5=10007.5, 75-4.93=70.07}.

[0073] Finally, the carbon capture process parameters are iteratively optimized to obtain the optimal carbon capture process parameters with the highest carbon capture score, which are then used to control the oxy-fuel combustion unit and the calcium cycle capture unit. The second carbon capture process parameters are then fed into S200 and S300, undergoing the same prediction-evaluation process to calculate the second carbon capture score. This iterative process continues, constantly comparing the old and new scores, retaining the better ones, and dynamically adjusting the step size. After several iterations, an optimal point is reached where the carbon capture score cannot be further improved by fine-tuning the parameters; these parameters are the optimal carbon capture process parameters. For example, the final optimal solution found is {10200 Nm³ / h, 72 t / h}, with a predicted carbon capture rate of 86%, a predicted impact on production of -0.2 tons, and a comprehensive score of 0.915. This optimal parameter set is then distributed to the oxy-fuel combustion unit and the calcium cycle capture unit for control. For example, the oxygen flow rate is set to 10200 Nm³ / h, and the adsorbent circulation rate is set to 72 t / h.

[0074] In this embodiment, a comprehensive scoring method is used to find an optimal balance between carbon capture efficiency and production guarantee, thereby maximizing the overall benefits for the enterprise. The adaptive step size adjustment mechanism enables the algorithm to quickly locate the optimal solution in the early stages and finely adjust it in the later stages to approach the optimal solution, improving the accuracy and efficiency of the optimization process, avoiding blind searching, and reducing the number of iterations and time required to reach the optimal state. From sensing demand, predicting effects, assessing risks, to decision optimization and final execution, the entire process requires no manual intervention and can automatically and continuously keep the carbon capture platform in the optimal operating state and dynamically respond to changes in production demand.

[0075] The embodiments of this application, through the specific implementation methods described above, achieve the following technical effects:

[0076] The process dynamic optimization method based on a carbon capture platform proposed in this application effectively solves the core problem of the difficulty in synergistically optimizing carbon capture efficiency and cement production stability in existing technologies. It achieves dynamic synergy between carbon capture and cement production, maximizing capture efficiency while ensuring production targets. It improves the intelligence and efficiency of the optimization process, taking into account both search speed and accuracy through an adaptive mechanism. It constructs a complete autonomous decision-making and control closed loop, which can respond to changes in operating conditions in real time and continuously maintain optimal operating conditions, providing a reliable technical path for the low-carbon transformation of energy-intensive industries.

[0077] Example 2, as Figure 2 As shown, this application provides a process dynamic optimization system based on a carbon capture platform, the system comprising:

[0078] The parameter acquisition module 11 is used to acquire the required output of cement production and the carbon capture process parameter space, wherein each carbon capture process parameter includes the oxygen combustion parameter and the calcium cycle parameter.

[0079] The parameter prediction module 12 is used to configure a first adjustment step size according to the required output, adjust and generate a first carbon capture process parameter within the carbon capture process parameter space, and predict and output a first predicted carbon capture parameter and a first calciner temperature in combination with the required output.

[0080] The production impact prediction module 13 is used to predict the impact on cement production based on the temperature of the first calcining furnace and the required output, and to obtain the first predicted impact output.

[0081] The dynamic optimization control module 14 is used to calculate a first carbon capture score based on the first predicted carbon capture parameters and the first predicted impact on the product, and adjust the first adjustment step size based on the first predicted impact on the output to obtain a second adjustment step size, and continue to iteratively optimize the carbon capture process parameters to obtain the optimal carbon capture process parameters, and control the all-oxygen combustion unit and the calcium cycle capture unit.

[0082] In one embodiment, the parameter acquisition module 11 is further configured to:

[0083] Obtain the current required output for cement production;

[0084] The oxygen flow rate adjustment space for all-oxygen combustion and the adsorbent circulation rate space for calcium-cycled carbon capture are obtained. These are combined to obtain the carbon capture process parameter space, where each carbon capture process parameter includes all-oxygen combustion parameters and calcium circulation parameters.

[0085] In one embodiment, the parameter prediction module 12 is further configured to:

[0086] Get the preset adjustment step size;

[0087] The ratio of the demand output to the average demand output over a historical period is calculated and used as the product adjustment coefficient. The preset adjustment step size is then adjusted to obtain the first adjustment step size.

[0088] Obtain the current carbon capture process parameters, adjust them using the first adjustment step size, and obtain the first carbon capture process parameters;

[0089] The first carbon capture process parameters and required output are input into the carbon capture predictor, and the first predicted carbon capture parameters and the first calciner temperature are output.

[0090] The steps for building the carbon capture predictor include:

[0091] Based on historical operational data of carbon capture in cement production, a set of sample carbon capture process parameters and a set of sample required output were collected. The carbon capture rate and calciner temperature under different sample carbon capture process parameters and sample required output were also collected. The sample predicted carbon capture parameter set and sample calciner temperature set were then labeled and obtained.

[0092] Based on machine learning, a trapping predictor is constructed, wherein the trapping predictor includes a trapping parameter prediction branch and a calcining furnace temperature prediction branch;

[0093] Using the sample carbon capture process parameter set and sample demand output set as input training data, and the sample predicted carbon capture parameter set and sample calciner temperature set as output supervision data, the capture predictor is trained and optimized until the test converges, thus completing the setup.

[0094] In one embodiment, the production impact prediction module 13 is further configured to:

[0095] Based on historical cement production data, we collected sample calciner temperature sets and sample demand output sets, and collected the output decreases when different sample calciner temperatures and sample demand outputs occurred, and labeled them to obtain the sample predicted output impact set.

[0096] A cement production impact predictor was constructed based on machine learning.

[0097] The cement production impact predictor is trained and optimized by using the sample calciner temperature set and the sample demand output set as input training data and the sample predicted impact output set as output supervision data until the test converges, thus completing the setup.

[0098] The temperature of the first calcining furnace and the required output are input into the cement production impact predictor, which then outputs the first predicted impact output.

[0099] In one embodiment, the dynamic optimization control module 14 is further configured to:

[0100] Based on the first prediction of the impact on product and demand output, the first actual output is calculated, and the first output achievement rate is calculated.

[0101] The first carbon capture score is calculated based on the first predicted carbon capture parameters and the first production achievement rate.

[0102] Specifically, based on the first predicted impact on production, the first adjustment step size is adjusted to obtain a second adjustment step size. The carbon capture process parameters are then iteratively optimized to obtain the optimal carbon capture process parameters, including:

[0103] Based on the first prediction of the impact on output and demand output, the step size adjustment coefficient is calculated, and the first adjustment step size is adjusted to obtain the second adjustment step size.

[0104] The carbon capture process parameters are adjusted using the second adjustment step size to obtain the second carbon capture process parameters.

[0105] Continue iterative optimization of carbon capture process parameters to obtain the optimal carbon capture process parameters with the highest carbon capture score, and control the all-oxygen combustion unit and the calcium cycle capture unit.

[0106] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0107] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0108] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. A process dynamic optimization method based on a carbon capture platform, characterized in that, The carbon capture platform includes an all-oxygen combustion unit and a calcium cycle capture unit, and the method includes: Obtain the current cement production demand and the carbon capture process parameter space, where each carbon capture process parameter includes the all-oxygen combustion parameter and the calcium cycle parameter; Based on the required output, configure the first adjustment step size, adjust and generate the first carbon capture process parameters within the carbon capture process parameter space, and combine the required output to predict and output the first predicted carbon capture parameters and the first calciner temperature. Based on the temperature and required output of the first calcining furnace, the impact on cement production is predicted, and the first predicted impact output is obtained. The first carbon capture score is calculated based on the first predicted carbon capture parameters and the first predicted impact on production. The first adjustment step size is adjusted based on the first predicted impact on production to obtain the second adjustment step size. The carbon capture process parameters are then iteratively optimized to obtain the optimal carbon capture process parameters, and the oxygen-fuel combustion unit and the calcium cycle capture unit are controlled.

2. The process dynamic optimization method based on a carbon capture platform according to claim 1, characterized in that, Obtain the current required output for cement production and the carbon capture process parameter space, including: Obtain the current required output for cement production; The oxygen flow rate adjustment space for all-oxygen combustion and the adsorbent circulation rate space for calcium-cycled carbon capture are obtained. These are combined to obtain the carbon capture process parameter space, where each carbon capture process parameter includes all-oxygen combustion parameters and calcium circulation parameters.

3. The process dynamic optimization method based on a carbon capture platform according to claim 1, characterized in that, Based on the required output, a first adjustment step size is configured, and first carbon capture process parameters are generated within the carbon capture process parameter space. Combined with the required output, first predicted carbon capture parameters and a first calcining furnace temperature are predicted and output, including: Get the preset adjustment step size; The ratio of the demand output to the average demand output over a historical period is calculated and used as the output adjustment coefficient. The preset adjustment step size is then adjusted to obtain the first adjustment step size. Obtain the current carbon capture process parameters, adjust them using the first adjustment step size, and obtain the first carbon capture process parameters; The first carbon capture process parameters and required output are input into the carbon capture predictor, and the first predicted carbon capture parameters and the first calciner temperature are output.

4. The process dynamic optimization method based on a carbon capture platform according to claim 3, characterized in that, The steps for building the carbon capture predictor include: Based on historical operational data of carbon capture in cement production, a set of sample carbon capture process parameters and a set of sample required output were collected. The carbon capture rate and calciner temperature under different sample carbon capture process parameters and sample required output were also collected. The sample predicted carbon capture parameter set and sample calciner temperature set were then labeled and obtained. Based on machine learning, a trapping predictor is constructed, wherein the trapping predictor includes a trapping parameter prediction branch and a calcining furnace temperature prediction branch; Using the sample carbon capture process parameter set and sample demand output set as input training data, and the sample predicted carbon capture parameter set and sample calciner temperature set as output supervision data, the capture predictor is trained and optimized until the test converges, thus completing the setup.

5. The process dynamic optimization method based on a carbon capture platform according to claim 1, characterized in that, Based on the temperature and required output of the first calcining furnace, the impact on cement production is predicted to obtain the first predicted impact on output, including: Based on historical cement production data, we collected sample calciner temperature sets and sample demand output sets, and collected the output decreases when different sample calciner temperatures and sample demand outputs occurred, and labeled them to obtain the sample predicted output impact set. A cement production impact predictor was constructed based on machine learning. The cement production impact predictor is trained and tuned using the sample calciner temperature set and the sample demand output set as input training data, and the sample predicted impact output set as output supervision data, until the test converges, thus completing the setup. The temperature of the first calcining furnace and the required output are input into the cement production impact predictor, which then outputs the first predicted impact output.

6. The process dynamic optimization method based on a carbon capture platform according to claim 1, characterized in that, A first carbon capture score is calculated based on the first predicted carbon capture parameters and the first predicted impact on production, including: Based on the first predicted impact on output and demand output, the first actual output is calculated, and the first output achievement rate is calculated. The first carbon capture score is calculated based on the first predicted carbon capture parameters and the first production achievement rate.

7. The process dynamic optimization method based on a carbon capture platform according to claim 1, characterized in that, Based on the first predicted impact on output, the first adjustment step size is adjusted to obtain a second adjustment step size. The carbon capture process parameters are then iteratively optimized to obtain the optimal carbon capture process parameters, including: Based on the first prediction of the impact on output and demand output, the step size adjustment coefficient is calculated, and the first adjustment step size is adjusted to obtain the second adjustment step size. The carbon capture process parameters are adjusted using the second adjustment step size to obtain the second carbon capture process parameters. Continue iterative optimization of carbon capture process parameters to obtain the optimal carbon capture process parameters with the highest carbon capture score, and control the all-oxygen combustion unit and the calcium cycle capture unit.

8. A process dynamic optimization system based on a carbon capture platform, characterized in that, The system is used to implement the process dynamic optimization method based on a carbon capture platform as described in any one of claims 1-7, the system comprising: The parameter acquisition module is used to obtain the required output for current cement production and to obtain the carbon capture process parameter space, wherein each carbon capture process parameter includes the all-oxygen combustion parameter and the calcium cycle parameter. The parameter prediction module is used to configure a first adjustment step size according to the demand output, adjust and generate the first carbon capture process parameters within the carbon capture process parameter space, and predict and output the first predicted carbon capture parameters and the first calciner temperature in combination with the demand output. The production impact prediction module is used to predict the impact on cement production based on the temperature of the first calcining furnace and the required output, and to obtain the first predicted impact output. The dynamic optimization control module is used to calculate a first carbon capture score based on the first predicted carbon capture parameters and the first predicted impact on production, and adjust the first adjustment step size based on the first predicted impact on production to obtain a second adjustment step size, and continue to iteratively optimize the carbon capture process parameters to obtain the optimal carbon capture process parameters, thereby controlling the all-oxygen combustion unit and the calcium cycle capture unit.

Images