Carbide slag desulfurization system parameter optimization method, system and equipment and storage medium

Abstract

The invention discloses a carbide slag desulfurization system parameter optimization method, system and equipment and a storage medium, and relates to the technical field of flue gas desulfurization, the method comprises the following steps: collecting parameters in a carbide slag desulfurization process in real time, and adjusting the collection frequency according to working conditions to obtain real-time data; based on the real-time data and the historical operation data, a fusion machine learning model is established, and after training and optimization, the optimal range of parameters of the desulfurization system is output; based on the optimal range of the parameters of the desulfurization system, the parameters of the desulfurization system are accurately adjusted; when the method is applied to a power plant, the ultrahigh desulfurization efficiency is realized, the operation cost is remarkably reduced, and the recovery rate and the purity of the byproduct gypsum are improved; the method can be widely applied to the fields of electric power, chemical engineering and the like, and has great economic and environmental benefits.

Description

Technical Field

[0001] This invention relates to the field of flue gas desulfurization technology, and in particular to a method, system, equipment and storage medium for optimizing parameters of a carbide slag desulfurization system. Background Technology

[0002] With increasing global emphasis on environmental protection, countries around the world have introduced stringent emission standards for air pollutants, strictly limiting emissions of sulfur dioxide (SO2) and other pollutants from industrial enterprises such as coal-fired power plants. To meet these environmental requirements, industrial enterprises urgently need efficient desulfurization technologies. Calcium carbide slag, as an industrial byproduct, mainly consists of CaO and Ca(OH)2 and possesses good desulfurization performance. Therefore, it is increasingly being used in flue gas desulfurization processes in coal-fired power plants, achieving both waste utilization and reduced desulfurization costs.

[0003] Before the advent of calcium carbide slag desulfurization technology, the limestone-gypsum method was the most widely used flue gas desulfurization technology. However, limestone mining causes environmental damage, and the transportation and processing costs of limestone are high. Furthermore, traditional desulfurization system parameters rely heavily on manual settings, lacking real-time optimization and dynamic adjustment capabilities, resulting in significant fluctuations in desulfurization efficiency and difficulty in consistently meeting stringent environmental standards. Therefore, finding a new desulfurizing agent and a more intelligent method for optimizing desulfurization system parameters has become a necessity for industry development. The composition and properties of calcium carbide slag are affected by various factors such as calcium carbide production processes and raw material quality, leading to unstable desulfurization performance. In addition, operating conditions such as SO2 concentration, temperature, and humidity in the flue gas also change over time and during the production process. Traditional desulfurization systems, lacking real-time monitoring and dynamic adjustment capabilities, cannot optimize desulfurization parameters in a timely manner based on these changes, resulting in significant fluctuations in desulfurization efficiency. During the calcium carbide slag desulfurization process, calcium ions (Ca2+) in the slurry... + Calcium sulfate (SO42-) reacts chemically with sulfate ions to form precipitates such as calcium sulfate (CaSO4). If the composition, temperature, pH, and other parameters of the slurry are not properly controlled, these precipitates can easily deposit on the inner walls, pipes, and nozzles of the desulfurization equipment, forming scale. Scale not only affects the normal operation of the equipment and reduces desulfurization efficiency, but also increases maintenance costs and energy consumption. The main byproduct of calcium carbide slag desulfurization is gypsum. However, due to the imperfect treatment and recovery technologies for byproducts in traditional desulfurization processes, the purity and quality of the gypsum byproduct are low, making it difficult to meet market demand. This results in low utilization of byproducts, wasting resources and increasing the processing costs for enterprises. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by this invention is the problem of fluctuating desulfurization efficiency, slurry scaling, low utilization rate of by-products, and lack of real-time optimization and dynamic adjustment capability of traditional desulfurization system parameters in the desulfurization process of carbide slag.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, embodiments of the present invention provide a method for optimizing parameters of a carbide slag desulfurization system, comprising:

[0008] Parameters during the desulfurization process of carbide slag are collected in real time, and the collection frequency is adjusted according to the operating conditions to obtain real-time data;

[0009] Based on real-time data and historical operating data, a fusion machine learning model is established, trained, and optimized to output the optimal range of desulfurization system parameters;

[0010] Based on the optimal range of desulfurization system parameters, the parameters of the desulfurization system are precisely adjusted.

[0011] As a preferred scheme for optimizing parameters of a carbide slag desulfurization system, the following is provided:

[0012] The real-time acquisition of parameters during the desulfurization process of carbide slag and the adjustment of the acquisition frequency according to the operating conditions include:

[0013] The rate of change of a specific parameter is used as the standard for judging the working condition. When the change of the specific parameter exceeds the first set threshold within a first set time, it is determined that the working condition is fluctuating violently, and the high-frequency acquisition mode is triggered.

[0014] The advantages of this preferred technical solution are: it can acquire more frequent data in a timely manner when the operating conditions fluctuate drastically, providing more accurate and timely information for subsequent parameter optimization, which is conducive to rapid response to unstable operating conditions and improves the adaptability and stability of the desulfurization system.

[0015] As a preferred scheme for optimizing parameters of a carbide slag desulfurization system, the following is provided:

[0016] The real-time acquisition of parameters during the desulfurization process of carbide slag, and the adjustment of the acquisition frequency according to the operating conditions, also includes:

[0017] When the variation of the specific parameter within the second set time period is less than the second set threshold, it is determined to be a stable operating condition, and the sampling period is relaxed.

[0018] The beneficial effects of this preferred technical solution are: under stable operating conditions, relaxing the sampling period can reduce the amount of data collected, reduce data processing pressure and cost, and at the same time meet the basic requirements for parameter monitoring under stable operating conditions.

[0019] As a preferred scheme for optimizing parameters of a carbide slag desulfurization system, the following is provided:

[0020] The process of establishing a fusion machine learning model based on real-time data and historical operational data, and then training and optimizing it, includes:

[0021] A model integrating multiple machine learning algorithms is constructed. Real-time data and historical operating data are standardized and then input into the fusion model. After training and optimization, the optimal range of slurry solid-liquid ratio, slurry pH control range, and slurry circulation flow control range are output.

[0022] The beneficial effects of this preferred technical solution are as follows: integrating multiple machine learning algorithms can give full play to the advantages of each algorithm, improve the accuracy and generalization ability of the model, and standardize the data processing to make the data more suitable for model training, thereby more accurately outputting the optimal range of desulfurization system parameters.

[0023] As a preferred scheme for optimizing parameters of a carbide slag desulfurization system, the following is provided:

[0024] The training and optimization include:

[0025] Supervised learning is used to train and optimize the model. The system input parameters and actual SO2 removal rate in the past running data are used as training samples. The gradient is updated using an optimizer, and the training effect is evaluated using a specific error function. The training set and validation set are divided, and cross-validation is performed after each round of training.

[0026] The beneficial effects of this preferred technical solution are as follows: supervised learning combined with appropriate training samples, gradient update methods and error evaluation functions enables the model to better learn the patterns in the data. Dividing the training set and validation set and performing cross-validation can effectively evaluate the model performance, avoid overfitting and improve the reliability of the model.

[0027] As a preferred scheme for optimizing parameters of a carbide slag desulfurization system, the following is provided:

[0028] The precise adjustment of desulfurization system parameters based on the optimal range of desulfurization system parameters includes:

[0029] Based on the difference between the current value of the parameter and the median of the prediction interval, a PID control strategy is used to gradually approach the target value. When the direction of change of the adjustment value remains consistent in the set number of cycles, the adjustment step size is automatically reduced to the set proportion of the initial value.

[0030] The beneficial effects of this preferred technical solution are as follows: the PID control strategy can adjust parameters more accurately, gradually approach the target value, and avoid parameter abrupt changes; automatically reducing the adjustment step size can prevent over-adjustment, making parameter adjustment more stable and accurate, and improving the operational stability of the desulfurization system.

[0031] As a preferred scheme for optimizing parameters of a carbide slag desulfurization system, the following is provided:

[0032] The precise adjustment of desulfurization system parameters based on the optimal range of desulfurization system parameters also includes:

[0033] An alarm mechanism for parameters exceeding limits is introduced. When the desulfurization efficiency or SO2 emission approaches the upper limit, the pH value is increased by the set unit and the slurry flow rate is increased by the set ratio.

[0034] The beneficial effects of this preferred technical solution are as follows: the parameter over-limit alarm mechanism can promptly detect abnormal situations in the operation of the desulfurization system. When the parameters are close to the upper limit, timely adjustment can quickly restore the desulfurization reaction rate, ensure the normal operation of the desulfurization system, and reduce the risk of pollutant emissions exceeding the standard.

[0035] Secondly, embodiments of the present invention provide a parameter optimization system for a carbide slag desulfurization system, comprising:

[0036] The real-time data acquisition module is used to collect parameters in the desulfurization process of carbide slag in real time, and adjust the acquisition frequency according to the working conditions to obtain real-time data.

[0037] The range determination module is used to establish a fusion machine learning model based on real-time data and historical operating data, and output the optimal range of desulfurization system parameters after training and optimization.

[0038] The adjustment module is used to precisely adjust the parameters of the desulfurization system based on the optimal range of the desulfurization system parameters.

[0039] Thirdly, embodiments of the present invention provide an electronic device, including:

[0040] Memory and processor;

[0041] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the one or more programs are executed by the one or more processors, the one or more processors implement the parameter optimization method for the carbide slag desulfurization system as described in any embodiment of the present invention.

[0042] Fourthly, embodiments of the present invention provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the aforementioned method for optimizing parameters of a carbide slag desulfurization system.

[0043] The beneficial effects of this invention are as follows: Through dynamic adjustment based on machine learning, the desulfurization efficiency is stabilized at over 99%, significantly higher than the 95%-98% of traditional desulfurization processes; slurry parameters are optimized, reducing equipment scaling, extending equipment lifespan, and lowering maintenance and cleaning costs; the introduction of machine learning algorithms allows the desulfurization system to automatically adjust operating parameters according to different working conditions through real-time monitoring and feedback adjustment, improving the desulfurization effect, and continuously optimizing and adapting to new operating conditions, reducing manual intervention and ensuring long-term stable operation of the system. Attached Figure Description

[0044] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0045] Figure 1 This is an overall flowchart of the parameter optimization method for the carbide slag desulfurization system described in this invention;

[0046] Figure 2 This is a schematic diagram of the machine learning model training and optimization process of the parameter optimization method for the carbide slag desulfurization system described in this invention. Detailed Implementation

[0047] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0048] Example 1, referring to Figure 1 This is the first embodiment of the present invention, which provides a method for optimizing parameters of a carbide slag desulfurization system, including:

[0049] S1: Real-time acquisition of parameters during the desulfurization process of carbide slag, and adjustment of acquisition frequency according to operating conditions to obtain real-time data;

[0050] S2: Based on real-time data and historical operating data, a fusion machine learning model is established, trained and optimized, and then the optimal range of desulfurization system parameters is output.

[0051] S3: Based on the optimal range of desulfurization system parameters, the parameters of the desulfurization system are precisely adjusted.

[0052] It should be noted that, through steps S1-S3, this embodiment integrates multiple real-time sensing parameters into a multi-model fusion algorithm and uses them to automatically generate an adjustable range; establishes a dynamic weighted fusion learning mechanism to achieve collaborative prediction of DNN / SVM / DT models, thereby improving the system's response robustness under operating condition fluctuations; constructs a closed-loop control path of "acquisition-judgment-prediction-adjustment-feedback", and combines an edge computing terminal and a PLC device to achieve a minute-level response speed.

[0053] Example 2, refer to Figures 1-2 As an embodiment of the present invention, based on the previous embodiment, a method for optimizing parameters of a carbide slag desulfurization system is provided, comprising:

[0054] In this embodiment, the real-time acquisition of parameters during the desulfurization process of carbide slag in step S1 above, and the adjustment of the acquisition frequency according to the operating conditions, includes:

[0055] Multiple sensors, such as flue gas SO2 concentration sensors, pH sensors, and slurry flow meters, are installed in the carbide slag desulfurization system to monitor important parameters in real time, including flue gas SO2 concentration, pH value in the desulfurization tower, slurry concentration, temperature, and flow rate.

[0056] The sampling frequency is adjusted according to changes in operating conditions. When the SO2 concentration in the flue gas changes drastically, the sampling frequency is increased to once every 30 seconds; while under stable operating conditions, the sampling frequency is reduced to once every 2 minutes to reduce the data processing pressure.

[0057] Specifically, the rate of change of SO2 concentration in flue gas is used as the criterion: if the SO2 concentration changes by more than ±20 mg / Nm³ within a unit time (60 seconds), the rate of change is considered significant. 3 If the operating conditions are considered to be fluctuating drastically, a high-frequency data acquisition mode will be automatically triggered (once every 30 seconds); if the SO2 concentration change is less than ±5 mg / Nm³ for 5 consecutive minutes... 3 If the condition is stable, the sampling period is relaxed to once every 2 minutes.

[0058] In another possible implementation, in addition to using the rate of change of SO2 concentration in the flue gas as the criterion, pressure changes within the desulfurization tower can also be considered. If the pressure fluctuation exceeds ±50 Pa within a unit time (60 seconds), it is considered a violent fluctuation in operating conditions, and the sampling frequency is increased to once every 30 seconds; if the pressure fluctuation is less than ±10 Pa for 5 consecutive minutes, it is considered a stable operating condition, and the sampling period is relaxed to once every 2 minutes. Simultaneously, changes in slurry temperature can also be considered; when the slurry temperature changes by more than ±3℃ in a short period, the sampling frequency is increased.

[0059] In this embodiment, step S2 above, which involves establishing a fusion machine learning model based on real-time data and historical operational data, and then training and optimizing it, includes:

[0060] Construct a machine learning model that integrates three algorithms: deep neural network (DNN), support vector machine (SVM), and decision tree (DT).

[0061] Specifically, the deep neural network structure has 4 layers, with 64, 128, 64 and 32 neurons in each layer, a learning rate of 0.001 and 1000 iterations; the support vector machine uses a radial basis function (RBF) kernel, with a penalty parameter C of 1.0 and a Gamma parameter of 0.01; the decision tree has a maximum depth of 10 and a minimum number of leaf nodes of 5.

[0062] After standardizing and processing real-time and historical operational data, the data is input into the fusion model. Through training and optimization, the optimal ranges for slurry solid-liquid ratio, slurry pH control, and slurry circulation flow rate are output.

[0063] Specifically, the standardization process includes mean removal, variance normalization, missing value imputation, and outlier removal. The Z-score method is used to transform all numerical parameters (such as SO2 concentration, pH, and temperature) to zero mean with unit variance; missing data is imputed using the moving average; and outliers are removed according to the 3σ rule. The processed data is stored in NumPy matrix format and used as the input feature matrix for the machine learning model.

[0064] like Figure 2 The diagram illustrates the machine learning model training and optimization process used in this embodiment. The model training employs supervised learning, using "system input parameters" and "actual SO2 removal rate" from past data as training samples. The Adam optimizer is used for gradient updates, and the training error function is the mean squared error (MSE). The training and validation sets are divided in an 8:2 ratio. After each training round, the model accuracy is cross-validated. After training, the learning rate, regularization coefficient, and model structure are fine-tuned using the GridSearch method to ultimately determine the fusion model structure and optimal parameters.

[0065] It should be noted that the fusion algorithm proposed in this embodiment is not a simple integration, but rather based on a weighted ensemble learning strategy. Weights are dynamically assigned according to the model's prediction accuracy on historical datasets (e.g., DNN 50%, SVM 30%, DT 20%), and the final decision result is output. Simultaneously, a model uncertainty metric (such as confidence interval fluctuation) is introduced to correct the confidence level of the prediction results, optimizing the robustness of the desulfurization parameter adjustment range. Compared to existing single-model schemes, this method improves generalization performance and stability under abnormal operating conditions.

[0066] In another possible implementation, in addition to the three algorithms—Deep Neural Network (DNN), Support Vector Machine (SVM), and Decision Tree (DT)—the Random Forest (RF) algorithm can also be introduced. The Random Forest algorithm has good resistance to overfitting and can handle high-dimensional data. Its parameters can be set as follows: 100 trees, a maximum number of features of sqrt, and a minimum number of sample splits of 2.

[0067] In another possible implementation, during model training, in addition to using the Adam optimizer, the Adagrad optimizer can be tried, which can adaptively adjust the learning rate for each parameter. Besides mean squared error (MSE), mean absolute error (MAE) can also be used as the training error function. During model fine-tuning, in addition to the GridSearch method, the Random Search method can be used, which can search a wider parameter space, increasing the probability of finding the optimal parameters.

[0068] In this embodiment, the precise adjustment of the desulfurization system parameters based on the optimal range of the desulfurization system parameters in step S3 above includes:

[0069] A real-time automatic adjustment system, namely a PLC control system, is constructed. Based on the optimal range of desulfurization system parameters output by the fusion model, the key process parameters of the slurry in the desulfurization tower are automatically adjusted.

[0070] Specifically, the automatic adjustment range for the slurry solid-liquid ratio is ±10% of the predicted range; the pH value of the slurry is adjusted by ±0.1 units each time; and the adjustment range for the slurry circulation flow rate is ±5% of the predicted range.

[0071] Furthermore, the automatic control system uses the difference between the current parameter value and the median of the predicted interval as a basis, and employs a PID control strategy to gradually approach the target value, avoiding abrupt adjustments. When the direction of adjustment change is consistent for three consecutive rounds (tending to be too high or too low), the system automatically reduces the adjustment step size to 50% of the initial value to avoid over-adjustment. In addition, the system supports a parameter over-limit alarm mechanism. Once the desulfurization efficiency or SO2 emission approaches the upper limit, the pH value is forcibly increased by 0.2 units and the slurry flow rate is increased by 10% to quickly restore the desulfurization reaction rate.

[0072] It should be noted that this step is based on the parameter prediction range output by the fusion model, and the automatic adjustment range is limited to this range. For example, if the predicted range of the slurry solid-liquid ratio is 1:3–1:5, then the control target range can be refined to 1:3.2–1:4.8; when the pH prediction range is 12.4–12.9, 12.5–12.8 is used as the stable adjustment window.

[0073] In another possible implementation, a fuzzy control strategy can be introduced based on the PID control strategy. The fuzzy control strategy can dynamically adjust control parameters according to different operating conditions and parameter changes, improving the flexibility and adaptability of the control. When the desulfurization efficiency or SO2 emissions approach the upper limit, in addition to forcibly increasing the pH value by 0.2 units and increasing the slurry flow rate by 10%, the amount of additives can also be increased to quickly restore the desulfurization reaction rate. Simultaneously, when the system detects that the content of a certain component in the slurry is too low, the corresponding substance is automatically replenished to ensure the normal progress of the desulfurization reaction.

[0074] Example 3: The above is an illustrative scheme of the parameter optimization method for the calcium carbide slag desulfurization system in this embodiment. It should be noted that the technical solution of the calcium carbide slag desulfurization system parameter optimization system and the technical solution of the above-described calcium carbide slag desulfurization system parameter optimization method belong to the same concept. Details not described in detail in the technical solution of the calcium carbide slag desulfurization system parameter optimization system in this embodiment can be found in the description of the technical solution of the above-described calcium carbide slag desulfurization system parameter optimization method.

[0075] This embodiment also provides a parameter optimization system for a carbide slag desulfurization system, including:

[0076] The real-time data acquisition module is used to collect parameters in the desulfurization process of carbide slag in real time, and adjust the acquisition frequency according to the working conditions to obtain real-time data.

[0077] The range determination module is used to establish a fusion machine learning model based on real-time data and historical operating data, and output the optimal range of desulfurization system parameters after training and optimization.

[0078] The adjustment module is used to precisely adjust the parameters of the desulfurization system based on the optimal range of the desulfurization system parameters.

[0079] This embodiment also provides an electronic device applicable to the parameter optimization method of carbide slag desulfurization system, including:

[0080] The system includes a memory and a processor. The memory stores computer-executable instructions, and the processor executes these instructions to implement the parameter optimization method for the carbide slag desulfurization system proposed in the above embodiments.

[0081] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the parameter optimization method for the carbide slag desulfurization system proposed in the above embodiments.

[0082] The storage medium proposed in this embodiment and the method for optimizing the parameters of the carbide slag desulfurization system proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0083] Example 4, referring to Table 1, is an embodiment of the present invention, providing a method for optimizing parameters of a carbide slag desulfurization system. To verify the beneficial effects of the present invention, a simulation experiment is conducted for scientific demonstration.

[0084] In a 300MW coal-fired power plant, a machine learning-based method for optimizing the parameters of a carbide slag desulfurization system was implemented. The experimental conditions are as follows:

[0085] Power plant capacity: 300MW;

[0086] SO2 concentration in flue gas: 2500 mg / Nm 3 ;

[0087] The solid-liquid ratio of calcium carbide slag slurry is 1:4.

[0088] Slurry circulation flow rate: 60m³ 3 / h.

[0089] Experimental steps:

[0090] 1. Install online monitoring devices: Install online monitoring devices such as pH sensors, SO2 concentration sensors, and flow meters in the carbide slag desulfurization reactor to collect parameters such as flue gas SO2 concentration, pH value in the desulfurization tower, slurry concentration, and flow rate in real time.

[0091] 2. Data Acquisition and Preprocessing: The acquired real-time data includes information such as SO2 concentration in flue gas, pH value of slurry in the desulfurization tower, and slurry flow rate. Data quality is ensured through data preprocessing techniques (such as noise reduction and standardization).

[0092] 3. Machine learning model training:

[0093] We collect operational data from the desulfurization system over a period of time, including parameters such as SO2 concentration, desulfurization efficiency, and slurry concentration under different operating conditions. Using this data, we train a machine learning model that can predict desulfurization efficiency under different parameters and generate optimal operating parameters.

[0094] 4. Real-time adjustment: The trained machine learning model is integrated into the control system to dynamically adjust operating parameters such as the solid-liquid ratio, pH value, and flow rate of the slurry based on real-time data and feedback signals. The adjustment process automatically optimizes system operation based on real-time changes in flue gas SO2 concentration.

[0095] 5. Feedback and Optimization: Based on the effects of the adjustments, data is collected through a feedback loop to continuously adjust the model, enabling the system to adapt to different operating conditions and environmental changes. The model is continuously optimized to ensure long-term, stable, and efficient desulfurization.

[0096] The experimental results are shown in Table 1:

[0097] Table 1. Experimental Results of Parameter Optimization for the Carbide Slag Desulfurization System

[0098] index Test Results <![CDATA[SO2 removal rate]]> 99.50％ purity of by-product gypsum 92.50％ Reduced system energy consumption 10％ Operating cost savings 500,000 yuan / year Scaling incidence 0.50％

[0099] After optimizing the desulfurization process parameters in real time using the method of this invention, the desulfurization efficiency has remained stable at over 99.5% for a long period, far exceeding the 90%-92% of the control group using traditional manual control; the SO2 emission concentration in the flue gas has been stably reduced to 20 mg / Nm³. 3 The following emission levels are significantly better than those achieved by traditional methods (50-100 mg / Nm³). 3 );

[0100] The method of this invention adopts real-time dynamic control, which increases the slurry utilization rate by 15% and reduces the overall energy consumption of the system by more than 10%; according to the actual operation data of the power plant, the annual operating cost is saved by more than 500,000 yuan.

[0101] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for optimizing parameters of a carbide slag desulfurization system, characterized in that, include: Parameters during the desulfurization process of carbide slag are collected in real time, and the collection frequency is adjusted according to the operating conditions to obtain real-time data; Based on real-time data and historical operating data, a fusion machine learning model is established, trained, and optimized to output the optimal range of desulfurization system parameters; Based on the optimal range of desulfurization system parameters, the parameters of the desulfurization system are precisely adjusted.

2. The method for optimizing parameters of a carbide slag desulfurization system as described in claim 1, characterized in that, The real-time acquisition of parameters during the desulfurization process of carbide slag and the adjustment of the acquisition frequency according to the operating conditions include: The rate of change of a specific parameter is used as the standard for judging the operating condition. When the change of the specific parameter exceeds the first set threshold within a first set time, it is determined that the operating condition is fluctuating drastically, and the high-frequency acquisition mode is triggered.

3. The method for optimizing parameters of a carbide slag desulfurization system as described in claim 2, characterized in that, The real-time acquisition of parameters during the desulfurization process of carbide slag, and the adjustment of the acquisition frequency according to the operating conditions, also includes: When the variation of the specific parameter within the second set time period is less than the second set threshold, it is determined to be a stable operating condition, and the sampling period is relaxed.

4. The method for optimizing parameters of a carbide slag desulfurization system as described in claim 3, characterized in that, The process of establishing a fusion machine learning model based on real-time data and historical operational data, and then training and optimizing it, includes: A model integrating multiple machine learning algorithms is constructed. Real-time data and historical operating data are standardized and then input into the fusion model. After training and optimization, the optimal range of slurry solid-liquid ratio, slurry pH control range, and slurry circulation flow control range are output.

5. The method for optimizing parameters of a carbide slag desulfurization system as described in claim 4, characterized in that, The training and optimization include: Supervised learning is used to train and optimize the model. The system input parameters and actual SO2 removal rate in the past running data are used as training samples. The gradient is updated using an optimizer, the training effect is evaluated using a specific error function, and the training set and validation set are divided. Cross-validation is performed after each round of training.

6. The method for optimizing parameters of a carbide slag desulfurization system as described in claim 5, characterized in that, The precise adjustment of desulfurization system parameters based on the optimal range of desulfurization system parameters includes: Based on the difference between the current value of the parameter and the median of the prediction interval, a PID control strategy is used to gradually approach the target value. When the direction of change of the adjustment value remains consistent in the set number of cycles, the adjustment step size is automatically reduced to the set proportion of the initial value.

7. The method for optimizing parameters of a carbide slag desulfurization system as described in claim 6, characterized in that, The precise adjustment of desulfurization system parameters based on the optimal range of desulfurization system parameters also includes: An alarm mechanism for parameters exceeding limits is introduced. When the desulfurization efficiency or SO2 emission approaches the upper limit, the pH value is increased by the set unit and the slurry flow rate is increased by the set ratio.

8. A parameter optimization system for a calcium carbide slag desulfurization system, using the method described in any one of claims 1 to 7, characterized in that, include: The real-time data acquisition module is used to collect parameters in the desulfurization process of carbide slag in real time, and adjust the acquisition frequency according to the operating conditions to obtain real-time data. The range determination module is used to establish a fusion machine learning model based on real-time data and historical operating data, and output the optimal range of desulfurization system parameters after training and optimization. The adjustment module is used to precisely adjust the parameters of the desulfurization system based on the optimal range of the desulfurization system parameters.

9. An electronic device, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions, which, when executed by the processor, implement the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the method according to any one of claims 1 to 7.

Images