Monitoring and emission control optimization adjustment method for sulfur trioxide

Abstract

The invention provides a sulfur trioxide monitoring and emission control optimization adjustment method, and belongs to the technical field of optimization adjustment, and the method comprises the steps: carrying out the real-time synchronous monitoring of sulfur trioxide in supercritical boiler flue gas through different types of sensors, carrying out the data preprocessing and data fusion processing of a synchronous monitoring result, and carrying out the real-time monitoring of the sulfur trioxide in the supercritical boiler flue gas; and correcting the processed data by combining a correction model established with the parameters of the supercritical boiler, carrying out deep mining on historical monitoring data, historical operation parameters of the supercritical boiler and historical operation states of the flue gas treatment equipment, establishing a sulfur trioxide control model, and carrying out real-time monitoring on the sulfur trioxide control model. And a cooperative control instruction of the flue gas treatment equipment and desulfurization adjustment parameters in a desulfurization system are determined according to an output result, optimization control adjustment is performed, a model is established, an effective control result is output through analysis of a current monitoring result, and optimization and regulation of the equipment are intelligently realized.

Description

Technical Field

[0001] The present invention relates to the technical field of optimization and adjustment, and particularly to a method for optimizing the monitoring and emission control of sulfur trioxide. Background Art

[0002] With the increasingly strict environmental protection requirements, the emission control of sulfur trioxide in the flue gas of supercritical boilers has become increasingly important. The emission of sulfur trioxide not only causes air pollution, leading to problems such as acid rain and haze that harm the environment and human health, but also may cause adverse effects such as corrosion to boilers and subsequent equipment. Currently, although there are some monitoring and control methods for sulfur trioxide in the flue gas of supercritical boilers, there are still many deficiencies in actual applications. For example, traditional monitoring methods may have problems such as insufficient accuracy and slow response speed, making it difficult to accurately reflect the emission situation in real time; existing emission control technologies often have poor coordination, unable to fully utilize the advantages of each flue gas treatment facility, resulting in low efficiency in removing sulfur trioxide.

[0003] Therefore, the present invention proposes a method for optimizing the monitoring and emission control of sulfur trioxide. Summary of the Invention

[0004] The present invention provides a method for optimizing the monitoring and emission control of sulfur trioxide, which is used to ensure the accuracy of sulfur trioxide concentration based on the synchronous monitoring of three types of sensors, and further provides a reliable basis for subsequent control through data preprocessing and fusion. By establishing a model and analyzing the current monitoring results, an effective control result is output, and the optimization and regulation of equipment are realized intelligently.

[0005] The present invention provides a method for optimizing the monitoring and emission control of sulfur trioxide, including: Step 1: Use different types of sensors to synchronously monitor sulfur trioxide in the flue gas of supercritical boilers in real time; Step 2: Perform data preprocessing and data fusion processing on the synchronous monitoring results, and correct the processed data in combination with a calibration model established based on supercritical boiler parameters; Step 3: Deeply excavate historical monitoring data, historical operating parameters of supercritical boilers, and historical operating states of flue gas treatment equipment to establish a sulfur trioxide control model; Step 4: Input the calibration result into the sulfur trioxide control model, and determine the coordinated control instructions for flue gas treatment equipment and the desulfurization adjustment parameters in the desulfurization system according to the output result for optimized control adjustment.

[0006] Preferably, deeply excavating historical monitoring data, historical operating parameters of supercritical boilers, and historical operating states of flue gas treatment equipment to establish a sulfur trioxide control model includes: Obtain the historical monitoring data of the supercritical boiler within a specified historical time period, where the historical monitoring information includes the sulfur trioxide concentration information at each time point within the historical specified time period; Divide the sulfur trioxide concentration information into multiple sub-informations based on a segmented time window, where the sub-information includes the concentration values of each measurement point involved at a specified time point; Match the historical operating parameters of the supercritical boiler and the historical operating status of the flue gas treatment equipment at each historical time point with the corresponding sub-information to obtain a first input sample; Obtain the sulfur trioxide concentration threshold of the supercritical boiler, the standard operating parameters of the supercritical boiler, and the theoretical working status of the flue gas treatment equipment at the concentration value of each history, and compare and analyze with the first input sample to obtain a first deviation sample; Determine the calibration coefficient of the first input sample and the first deviation sample under each sample element, and determine the regulation loss level according to the calibration coefficient and the element type of the corresponding sample element; Obtain the reference level of the corresponding sample element based on all the regulation loss levels under the same sample element, and sort the improvement directions of all sample elements in sequence to obtain a first output sample; Construct a sample set based on all the first input samples and the first output samples corresponding to each first input sample; Train and test the neural network model according to the sample set to obtain a regulation model.

[0007] Preferably, determining the regulation loss level includes:

[0008] where, represents the regulation loss level of the corresponding sample element; represents the measured value of the i-th sample element in the first input sample; represents the theoretical value of the i-th sample element; represents the maximum value of the i-th sample element regarding the measured value under all the first input samples; represents the minimum value of the i-th sample element regarding the measured value under all the first input samples; represents the preset regulation loss coefficient of the h-th element type; represents the calibration coefficient of the i-th sample element.

[0009] Preferably, training and testing the neural network model according to the sample set to obtain a regulation model includes: Determine the division ratio of the sample set, and divide the sample set based on the division ratio to obtain a training set and a test set; Train the neural network model according to the training set; Test the trained model according to the test set to obtain a regulation model.

[0010] Preferably, training the neural network model according to the training set includes: Receive each training sample in the training set based on the input layer of the neural network; Based on multiple convolutional layers of the neural network and the non-linear activation function of each convolutional layer, perform dimension extraction on each training sample received by the multi-input layer to determine the dimension control feature of each training sample; Based on multiple deep convolutional layers of the neural network, perform noise extraction on each training sample in the training set to determine the noise control feature of each training sample; Use multiple deconvolutional layers of the neural network to respectively reconstruct the dimension control feature and the noise control feature of each training sample in the training set to obtain a noise-free sample; Determine the difference value between each training sample and the corresponding noise-free sample. If all the difference values are less than the preset value, it is determined that the training of the neural network model is completed.

[0011] Preferably, determining the difference value between each training sample and the corresponding noise-free sample includes:

[0012]

[0013] Wherein, represents the difference value between the u-th training sample and the corresponding noise-free sample; represents the dimension loss value of the u-th training sample; , are respectively the values of the i-th sample element under the u-th training sample and the u-th noise-free sample; N1 represents the total number of sample elements existing in each training sample; represents the differential fine-tuning coefficient, with a value of 0.01; represents the expectation of.

[0014] Preferably, different types of sensors include: sensors based on the controlled condensation method, sensors based on ion chromatography, and sensors based on optical absorption; Different types of sensors appear in a combined deployment at each monitoring location.

[0015] Preferably, correcting the processed data in combination with a correction model established based on the supercritical boiler parameters includes: Collect the parameters of the supercritical boiler during its operation, where the parameters of the supercritical boiler include: boiler load, coal type, combustion temperature, and excess air coefficient; Apply the multiple regression analysis method to establish the initial mathematical relationship between the concentration value of sulfur trioxide and the parameters of the supercritical boiler, and use the least squares method to determine the correction coefficient between the concentration value of sulfur trioxide and each parameter of the supercritical boiler as the coefficient to be supplemented for the corresponding parameter of the supercritical boiler in the initial mathematical relationship, so as to obtain the correction model; Correct the concentration value in the processed data based on the correction model.

[0016] Compared with the prior art, the beneficial effects of the present application are as follows: Based on the synchronous monitoring of three sensors, the accuracy of the sulfur trioxide concentration is ensured, and further, through the preprocessing and fusion of the data, a reliable basis is provided for subsequent control. By establishing a model and analyzing the current monitoring results, an effective control result is output, and the optimization and regulation of the equipment are realized intelligently. Description of the Drawings

[0017] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0018] Figure 1 It is a flowchart of the method for monitoring and optimizing the emission control of sulfur trioxide in the embodiment of the present invention. Detailed Embodiments

[0019] To make the objectives, technical solutions, and advantages of the present invention clearer, the following will clearly and completely describe the technical solutions in the present invention with reference to the drawings in the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art without creative efforts based on the embodiments of the present invention belong to the scope of protection of the present invention.

[0020] The present invention provides a method for monitoring and optimizing the emission control of sulfur trioxide, as Figure 1 shown, including: Step 1: Use different types of sensors to synchronously monitor sulfur trioxide in the flue gas of the supercritical boiler in real time; Step 2: Perform data preprocessing and data fusion processing on the synchronous monitoring results, and combine with the correction model established with the parameters of the supercritical boiler to correct the processed data; Step 3: Conduct in-depth mining of historical monitoring data, historical operating parameters of supercritical boilers, and historical operating status of flue gas treatment equipment to establish a sulfur trioxide control model; Step 4: Input the correction results into the sulfur trioxide control model, determine the coordinated control instructions of the flue gas treatment equipment and the desulfurization adjustment parameters in the desulfurization system according to the output results, and perform optimized control adjustments.

[0021] In this embodiment, three sensors are used to synchronously measure the flue gas in the boiler, and finally the concentration is averaged to obtain the concentration value monitored at the corresponding position, and the sensors are: sensors under controlled condensation method, ion chromatography and optical absorption method.

[0022] In this embodiment, in the data preprocessing stage, the 3σ principle is used to identify and eliminate abnormal data, and linear interpolation is used to fill in the missing data; data fusion uses a weighted average fusion algorithm to assign weights according to the historical measurement accuracy of each sensor (such as controlling the condensation method sensor weight 0.4, the ion chromatography sensor weight 0.3, and the optical absorption method sensor weight 0.3).

[0023] The calibration model is pre-built.

[0024] In this embodiment, boiler operating parameters (such as load, temperature, pressure) and equipment operating status data (such as desulfurization tower pH value, dust removal efficiency, denitrification ammonia escape rate) are used as input, and equipment control parameters (such as desulfurization agent dosage, dust removal electric field voltage, denitrification ammonia injection amount) are used as output. In this embodiment, the corrected sulfur trioxide concentration data is input into the sulfur trioxide control model, and the model outputs the equipment collaborative control instructions and desulfurization adjustment parameters. The control instructions are transmitted to the PLC control system through the Modbus protocol to realize the operation of equipment such as the temperature adjustment of the air preheater, the flow control of the slurry circulation pump of the desulfurization tower, the injection frequency adjustment of the pulse valve of the dust removal system, and the opening adjustment of the ammonia injection grid of the denitrification system, so as to optimize the sulfur trioxide emission control in real time.

[0025] The beneficial effects of the above technical solution are: based on the synchronous monitoring of the three sensors, the accuracy of the sulfur trioxide concentration is guaranteed, and further through the preprocessing and fusion of the data, a reliable basis is provided for subsequent control, and by establishing a model and analyzing the current monitoring results to output effective control results, the optimization and regulation of the equipment are realized intelligently.

[0026] The present invention provides a method for monitoring and optimizing the emission control of sulfur trioxide, which deeply mines historical monitoring data, historical operating parameters of supercritical boilers, and historical operating status of flue gas treatment equipment to establish a sulfur trioxide control model, including: Obtain the historical monitoring data of the supercritical boiler within a specified historical time period, where the historical monitoring information includes the sulfur trioxide concentration information at each time point within the specified historical time period; Divide the sulfur trioxide concentration information into multiple sub-informations based on segmented time windows, where the sub-information includes the concentration values of each measurement point involved at a specified time point; Compare and match the historical operating parameters of the supercritical boiler and the historical operating status of the flue gas treatment equipment at each historical time point with the corresponding sub-information to obtain the first input sample; Obtain the sulfur trioxide concentration threshold of the supercritical boiler, the standard operating parameters of the supercritical boiler, and the theoretical working status of the flue gas treatment equipment at the concentration value of each history, and compare and analyze with the first input sample to obtain the first deviation sample; Determine the calibration coefficient of the first input sample and the first deviation sample under each sample element, and determine the regulation loss level according to the calibration coefficient and the element type of the corresponding sample element; Obtain the reference level of the corresponding sample element based on all the regulation loss levels under the same sample element, and sort the improvement directions of all sample elements in sequence to obtain the first output sample; Construct a sample set based on all the first input samples and the first output samples corresponding to each first input sample; Train and test the neural network model according to the sample set to obtain the regulation model.

[0027] Preferably, determining the regulation loss level includes:

[0028] Wherein, represents the regulation loss level of the corresponding sample element; represents the measured value of the i-th sample element in the first input sample; represents the theoretical value of the i-th sample element; represents the maximum value of the i-th sample element regarding the measured value under all the first input samples; represents the minimum value of the i-th sample element regarding the measured value under all the first input samples; represents the preset regulation loss coefficient of the h-th element type; represents the calibration coefficient of the i-th sample element.

[0029] In this embodiment, January 1, 2024 - January 31, 2024 was selected as the historical specified time period, and the sulfur trioxide concentration information at each 5 - minute time point within this time period was successfully obtained, with a total of 8784 data records. The sulfur trioxide concentration shows fluctuating changes in different time periods, with the lowest concentration being 8mg / m³ and the highest concentration reaching 18mg / m³, providing a rich data basis for subsequent analysis.

[0030] The continuous sulfur trioxide concentration information was divided according to a certain time interval (i.e., the segmented time window) to form multiple sub - information segments. Each sub - information contains the sulfur trioxide concentration values at each measurement point to be measured (such as the sensor installation points at different positions in the flue) at the specified time point. This division facilitates subsequent refined analysis and processing of the data, and explores the concentration change characteristics in different time periods. Taking 1 hour as the segmented time window, the sulfur trioxide concentration information obtained from the above experiment was divided, and a total of 448 sub - information segments were obtained. Through analysis, it was found that there are differences in the concentration values of each measurement point in different sub - information segments. The sulfur trioxide concentration difference between the upstream and downstream of the flue in some sub - information segments is relatively large, reflecting the diffusion and reaction of the flue gas in the flue, and providing effective data units for studying the concentration distribution law.

[0031] The historical operating parameters of the supercritical boiler (such as boiler load, combustion temperature, steam pressure, etc.) and the historical operating status of the flue gas treatment equipment (such as the pH value of the desulfurization tower, the electric field strength of the dust removal equipment, the ammonia injection volume of the denitration system, etc.) were associated and matched with the sulfur trioxide concentration sub - information at the corresponding time points. Ensure that each sub - information corresponds to the complete boiler operating conditions and equipment status data at the same time point, and combine them to form the first input sample, which serves as the basic data set for subsequent analysis.

[0032] During the matching process, each sub - information was successfully associated with the corresponding boiler operating parameters and equipment status data, forming 448 first input samples. For example, in a certain sub - information segment, the sulfur trioxide concentration is 12mg / m³, the corresponding boiler load is 500MW, the combustion temperature is 1300℃, the pH value of the desulfurization tower is 5.8, the electric field strength of the dust removal equipment is 35kV, etc. These data together constitute a complete first input sample, clearly presenting the operating conditions of the system at this time point.

[0033] Determine the environmental protection emission standards or process requirement thresholds for the sulfur trioxide concentration of the supercritical boiler, as well as the standard operating parameter ranges of the boiler and the ideal working state parameters of the flue gas treatment equipment at different sulfur trioxide concentrations. Then, compare the actual data in the first input sample with these standard or theoretical data, calculate the deviation values of each data item, and thus obtain the first deviation sample, which is used to reflect the degree of difference between the actual operating situation and the ideal state.

[0034] After comparative analysis, among 448 first input samples, it was found that in 87 samples, the sulfur trioxide concentration exceeded the threshold, some boiler operating parameters deviated from the standard range, and the operating status of the flue gas treatment equipment did not reach the theoretical optimal state. For example, in a certain sample, the sulfur trioxide concentration was 16 mg / m³ (the threshold was 15 mg / m³), the boiler load exceeded the standard range by 10%, and the pH value of the desulfurization tower was 0.3 lower than the theoretical value. These deviation data were recorded to form the first deviation sample, visually showing the problem points during operation.

[0035] The return-to-standard coefficient is used to measure the cost or difficulty required to adjust each data element in the first input sample back to the ideal state after deviating from the standard or theoretical value. According to the nature of different data elements (such as different element types like sulfur trioxide concentration, boiler load, etc.), combined with the magnitude of the return-to-standard coefficient, the sample elements are classified into different regulation loss levels, such as low loss level, medium loss level, and high loss level, so as to formulate targeted regulation strategies subsequently.

[0036] In the comparison between the first input sample and the first deviation sample, the return-to-standard coefficient of each sample element was determined. For example, for the sulfur trioxide concentration element, if the return-to-standard coefficient is 0.8 (the larger the value, the greater the adjustment difficulty), according to its element type and the set classification standard, it is classified into the high loss level; while for some equipment operating status parameters, the return-to-standard coefficient is 0.3, which is classified into the low loss level. Finally, all sample elements are classified into different regulation loss levels, providing a quantitative basis for optimization and adjustment.

[0037] For the same type of sample elements (such as all sulfur trioxide concentration elements), by synthesizing their regulation loss levels in each first input sample, the reference level of this sample element is calculated to reflect the overall regulation difficulty level of this element. Then, according to the reference level, the improvement priorities of all sample elements are sorted to determine which elements to adjust first can more efficiently improve the system performance. The sorting results and relevant information are organized to form the first output sample. The reference level of the sulfur trioxide concentration element is high, the reference level of the boiler load element is medium, and the reference level of the pH value element of the desulfurization tower is low. According to the reference level, the order of the improvement directions is determined as: first adjust the parameters related to sulfur trioxide concentration, followed by the boiler load, and finally the pH value of the desulfurization tower and other parameters.

[0038] Each sample pair (the first input sample and the first output sample) represents an actual operating state and its corresponding optimization direction. The sample set covers various operating conditions and optimization strategies in historical operations, providing sufficient data support for the subsequent training of the neural network model. The sample set includes both optimization schemes under high load and high sulfur trioxide emission conditions, as well as adjustment strategies under low load and low concentration emissions, providing rich data resources for the model to learn complex operating laws and optimization logics.

[0039] Neural networks such as BP neural networks, LSTM neural networks, etc.

[0040] The beneficial effects of the above technical solution are as follows: Calculate the deviation value of each data item to obtain the first deviation sample, which is used to reflect the degree of difference between the actual operation situation and the ideal state. Classify all sample elements into different regulation loss levels, providing a quantitative basis for optimization and adjustment. The rich samples provide rich data resources for the model to learn complex operation rules and optimization logics.

[0041] The present invention provides a method for monitoring sulfur trioxide and optimizing and adjusting emission control. Train and test a neural network model according to the sample set to obtain a regulation model, including: Determine the division ratio of the sample set, and divide the sample set based on the division ratio to obtain a training set and a test set; Train the neural network model according to the training set; Test the trained model according to the test set to obtain a regulation model.

[0042] The beneficial effects of the above technical solution are as follows: Effectively ensure the training and testing of the model by proportionally dividing the sample set.

[0043] The present invention provides a method for monitoring sulfur trioxide and optimizing and adjusting emission control. Train a neural network model according to the training set, including: Receive each training sample in the training set based on the input layer of the neural network; Perform dimension extraction on each training sample received by the multi-input layer based on multiple convolutional layers of the neural network and the non-linear activation function of each convolutional layer to determine the dimension control features of each training sample; Perform noise extraction on each training sample in the training set based on multiple deep convolutional layers of the neural network to determine the noise control features of each training sample; Use multiple deconvolutional layers of the neural network to respectively perform feature reconstruction on the dimension control features and noise control features of each training sample in the training set to obtain noise-free samples; Determine the difference value between each training sample and the corresponding noise-free sample. If all difference values are less than a preset value, it is determined that the training of the neural network model is completed.

[0044] Preferably, determining the difference value between each training sample and the corresponding noise-free sample includes:

[0045]

[0046] Among them, It represents the difference value between the u-th training sample and the corresponding noise-free sample; It represents the dimensional loss value of the u-th training sample; 、 They are respectively the values of the i-th sample element under the u-th training sample and the u-th noise-free sample; N1 represents the total number of sample elements existing in each training sample; It represents the differential fine-tuning coefficient, and its value is 0.01; It represents the expectation of

[0047] In this embodiment, the input layer is the interface for the neural network to interact with external data. Its function is to convert each training sample in the training set (in the scenario of controlling sulfur trioxide emissions in a supercritical boiler, the training sample is the sample pair composed of the previously constructed first input sample and the first output sample) into a numerical form that the neural network can process and transmit it to the subsequent network layer. The number of neurons in the input layer is usually consistent with the number of features of the training sample. For example, if the training sample contains 10 features such as boiler load, sulfur trioxide concentration, and pH value of the desulfurization tower, then the input layer has 10 neurons, and each neuron corresponds to a feature value. In the training experiment of the supercritical boiler sulfur trioxide emission control model, the training set contains 358 groups of sample pairs, and the input layer successfully receives and converts all training samples. Taking one group of samples as an example, the input layer accurately transmits the feature values such as boiler load of 520 MW, sulfur trioxide concentration of 14 mg / m³, and pH value of the desulfurization tower of 5.6 in this sample to the subsequent network layer.

[0048] After being processed by 3 convolutional layers and the ReLU activation function, dimensional control features are successfully extracted from the training samples. For example, the training samples originally containing multiple complex features are processed to obtain dimensional control features such as the change trend feature of the boiler load and the fluctuation pattern feature of the sulfur trioxide concentration. Through visual analysis, it is found that these features can clearly show the internal structure of the data in different dimensions, are more representative than the original data, and provide more effective information for the subsequent model learning.

[0049] Using 5 deep convolutional layers to process the training samples, noise control features are successfully extracted. Through analysis, it is found that the noise in some training samples mainly comes from abnormal data fluctuations caused by sensor measurement errors and interference during data transmission. For example, in some samples, there are short abnormal jumps in the sulfur trioxide concentration, and the deep convolutional layer identifies this abnormal change as noise and forms corresponding noise control features.

[0050] After reconstructing the dimension control features and noise control features through two deconvolution layers, a noise-free sample is obtained. By comparing the original training samples and the noise-free samples, it is found that the data in the noise-free samples is smoother and more reasonable, removing the abnormal fluctuations and noise interference in the original samples. For example, in a certain sample, the original sulfur trioxide concentration data had severe fluctuations caused by noise. After being processed by the deconvolution layer, the changing trend of the concentration data is more in line with the actual operation law, providing higher-quality data for model training.

[0051] During the training process, the mean square error is set as the measurement method for the difference value, and the preset value is 0.01. As the training progresses, the mean square error between the training samples and the noise-free samples is continuously calculated. When training reaches the 45th epoch, the mean square error of all training samples is less than 0.01, meeting the preset conditions, and it is determined that the neural network model training is completed.

[0052] The preset value is a threshold preset during the model training process and serves as a standard for judging whether the model training is completed, which is reasonably set according to the actual application scenarios and requirements.

[0053] The beneficial effects of the above technical solution are as follows: The noise-free samples can more accurately reflect the true characteristics and laws of the data compared with the original samples, which helps to improve the model training effect. Through the verification of the test set, the model also shows good prediction performance on the test set, proving that the trained model has good generalization ability. By comparing the difference value between the training samples and the noise-free samples with the preset value, it is judged whether the neural network model has reached sufficient training accuracy and performance. When the conditions are met, it is determined that the model training is completed and can be put into use.

[0054] The present invention provides a method for monitoring sulfur trioxide and optimizing and adjusting emission control. Different types of sensors include: sensors based on the controlled condensation method, sensors based on ion chromatography, and sensors based on optical absorption method; Different types of sensors are deployed in combination at each monitoring location.

[0055] The beneficial effects of the above technical solution are as follows: By synchronously deploying three types of sensors, it is convenient to ensure the accuracy of measurement.

[0056] The present invention provides a method for monitoring sulfur trioxide and optimizing and adjusting emission control. Combining with the calibration model established based on the supercritical boiler parameters, the processed data is calibrated, including: Collecting the supercritical boiler parameters during the operation of the supercritical boiler, wherein the supercritical boiler parameters include: boiler load, coal type, combustion temperature, and excess air coefficient; Using the multiple regression analysis method, an initial mathematical relationship between the concentration value of sulfur trioxide and the supercritical boiler parameters is established. The least squares method is used to determine the correction coefficients between the concentration value of sulfur trioxide and each supercritical boiler parameter, which are used as the coefficients to be supplemented for the corresponding supercritical boiler parameters in the initial mathematical relationship, and a correction model is obtained. Based on the correction model, the concentration values in the processed data are corrected.

[0057] In this embodiment, relevant data are collected in real time through sensors, instruments and coal quality analysis equipment installed in various parts of the boiler. For example, the boiler load can be calculated by monitoring parameters such as steam flow and pressure; the coal type is determined through coal quality analysis to obtain its sulfur content, volatile matter and other characteristics; the combustion temperature is measured by temperature sensors such as thermocouples in the furnace; and the excess air coefficient is calculated based on flue gas composition analysis data (such as oxygen content). These parameters will serve as the basic variables for establishing the subsequent mathematical relationship.

[0058] In a 72-hour continuous operation monitoring experiment of a 600MW supercritical boiler, complete supercritical boiler parameter data are successfully collected. Among them, the boiler load fluctuates between 400 - 600MW, involving 3 different coal types (low-sulfur coal, medium-sulfur coal, high-sulfur coal respectively), the combustion temperature range is 1200 - 1400°C, and the excess air coefficient varies between 1.1 - 1.3. These data provide a rich sample for subsequent analysis.

[0059] The initial mathematical relationship is: ye = b0 + b1f1 + b2f2 + b3f3 + b4f4 + p0, where ye represents the concentration value, f1, f2, f3, f4 represent the boiler load, coal type, combustion temperature, excess air coefficient respectively, b0 is the constant term; b1 to b4 are the correction coefficients to be determined, p0 is the random error term, and the obtained correction model is: ye = -5.2 + 0.012f1 + 2.5f2 + 0.003f3 - 1.8f4 + 0.01.

[0060] Taking one time point as an example, the sulfur trioxide concentration monitoring data are corrected. The original monitored concentration is 12mg / m³. At this time, the boiler load is 550MW, the coal type is high-sulfur coal (coded 3), the combustion temperature is 1350°C, and the excess air coefficient is 1.2. Substituting into the correction model, the corrected concentration value is calculated to be 13.8mg / m³. By comparing with the laboratory offline analysis data, the relative error between the corrected data and the actual concentration is reduced from the original 15% to 5%, significantly improving the accuracy of the sulfur trioxide concentration monitoring data.

[0061] The beneficial effects of the above technical solution are as follows: By establishing a mathematical relationship, the relationship between the parameters of a supercritical boiler and the sulfur trioxide concentration value can be quantitatively described, and the original concentration monitoring value can be adjusted to eliminate the measurement error caused by the change of boiler operation parameters, so as to obtain corrected data that more accurately reflects the actual sulfur trioxide concentration.

[0062] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for monitoring sulfur trioxide and optimizing emission control adjustments, characterized in that, Including: Step 1: Use different types of sensors to perform real-time synchronous monitoring of sulfur trioxide in the flue gas of a supercritical boiler; Step 2: Conduct data preprocessing and data fusion processing on the synchronous monitoring results, and combine with a calibration model established for supercritical boiler parameters to calibrate the processed data; Step 3: Deeply mine the historical monitoring data, historical operating parameters of the supercritical boiler, and historical operating status of the flue gas treatment equipment to establish a sulfur trioxide control model; Step 4: Input the calibration result into the sulfur trioxide control model, determine the collaborative control instructions for the flue gas treatment equipment and the desulfurization adjustment parameters in the desulfurization system according to the output result, and perform optimized control adjustment.

2. The method for monitoring sulfur trioxide and optimizing and adjusting emission control according to claim 1, wherein Deeply mining the historical monitoring data, historical operating parameters of the supercritical boiler, and historical operating status of the flue gas treatment equipment to establish a sulfur trioxide control model, including: Obtain the historical monitoring data of the supercritical boiler within a historical specified time period, where the historical monitoring information includes the sulfur trioxide concentration information at each time point within the historical specified time period; Divide the sulfur trioxide concentration information into multiple sub-informations based on a segmented time window, where the sub-information includes the concentration values of each measurement point involved at a specified time point; Match the historical operating parameters of the supercritical boiler and the historical operating status of the flue gas treatment equipment at each historical time point with the corresponding sub-information to obtain a first input sample; Obtain the sulfur trioxide concentration threshold of the supercritical boiler, the standard operating parameters of the supercritical boiler, and the theoretical working status of the flue gas treatment equipment at each historical concentration value, and compare and analyze with the first input sample to obtain a first deviation sample; Determine the regression coefficient of the first input sample and the first deviation sample under each sample element, and determine the regulation loss level according to the regression coefficient and the element type of the corresponding sample element; Obtain the reference level of the corresponding sample element based on all the regulation loss levels under the same sample element, and sort the improvement directions of all sample elements in sequence to obtain a first output sample; Construct a sample set based on all the first input samples and the first output samples corresponding to each first input sample; Train and test the neural network model according to the sample set to obtain a regulation model.

3. The monitoring and emission control optimization adjustment method for sulfur trioxide according to claim 2, wherein Determine the regulation loss level, including: Among them, represents the regulation loss level corresponding to the sample element; represents the measured value of the i-th sample element in the first input sample; represents the theoretical value of the i-th sample element; represents the maximum value of the measured values of the i-th sample element under all first input samples; represents the minimum value of the measured values of the i-th sample element under all first input samples; represents the preset regulation loss coefficient of the h-th element type; represents the re-calibration coefficient of the i-th sample element.

4. The monitoring and emission control optimization adjustment method for sulfur trioxide according to claim 3, wherein Training and testing the neural network model according to the sample set to obtain a regulation model, including: Determine the division ratio of the sample set, and divide the sample set based on the division ratio to obtain a training set and a test set; Train the neural network model according to the training set; Test the trained model according to the test set to obtain a regulation model.

5. The monitoring and emission control optimization adjustment method for sulfur trioxide according to claim 4, characterized in that Training the neural network model according to the training set, including: Receive each training sample in the training set based on the input layer of the neural network; Extract the dimensions of each training sample received by the input layer through multiple convolutional layers of the neural network and the non-linear activation function of each convolutional layer to determine the dimension control characteristics of each training sample; Multiple deep convolutional layers based on a neural network extract noise from each training sample in the training set to determine the noise control features of each training sample; Use multiple deconvolutional layers of the neural network to respectively reconstruct the dimensional control features and noise control features of each training sample in the training set to obtain noise-free samples; Determine the difference value between each training sample and the corresponding noise-free sample. If all the difference values are less than a preset value, it is determined that the training of the neural network model is completed.

6. The monitoring and emission control optimization adjustment method for sulfur trioxide according to claim 4, wherein Determining the difference value between each training sample and the corresponding noise-free sample includes: Among them, represents the difference value between the u-th training sample and the corresponding noise-free sample; represents the dimensional loss value of the u-th training sample; , are the values of the i-th sample element under the u-th training sample and the u-th noise-free sample respectively; N1 represents the total number of sample elements existing in each training sample; represents the differential fine-tuning coefficient, with a value of 0.01; represents the expectation of.

7. The monitoring and emission control optimization adjustment method for sulfur trioxide according to claim 1, wherein Different types of sensors include: sensors based on the controlled condensation method, ion chromatography-based sensors, and optical absorption-based sensors; Different types of sensors appear in a combined deployment at each monitoring location.

8. The method for monitoring sulfur trioxide and optimizing and adjusting emission control according to claim 1, characterized in that Combined with the calibration model established with the supercritical boiler parameters, calibrate the processed data, including: Collect the supercritical boiler parameters during the operation of the supercritical boiler, where the supercritical boiler parameters include: boiler load, coal type, combustion temperature, and excess air coefficient; Apply the multiple regression analysis method to establish the initial mathematical relationship between the concentration value of sulfur trioxide and the supercritical boiler parameters, and use the least squares method to determine the calibration coefficient between the concentration value of sulfur trioxide and each supercritical boiler parameter as the coefficient to be supplemented for the corresponding supercritical boiler parameter in the initial mathematical relationship to obtain the calibration model; Calibrate the concentration value in the processed data based on the calibration model.

Images