Flue gas desulfurization control system and method

Abstract

The invention discloses a flue gas desulfurization control system and method, and relates to the technical field of flue gas desulfurization control, and the system comprises a pipeline module which is used for collecting flue gas data and slurry data, a spray tower module which is internally provided with an ultrasonic generator and is used for collecting the slurry coverage rate, and a processing module. The pipeline module is used for collecting a historical training data set in advance, training a machine learning model for predicting the slurry coverage rate in the spray tower module based on the historical training data set, and collecting flue gas data and slurry data in real time; the processing module predicts the slurry coverage rate in the spray tower module by using the machine learning module based on the flue gas data and the slurry data collected in real time, and adjusts the output power of the ultrasonic generator based on the predicted slurry coverage rate; the ultrasonic generator is used for carrying out ultrasonic treatment on mist liquid formed by slurry in the inner space of the spray tower, distribution of the mist liquid is adjusted, the coverage rate of the slurry is matched with desulfurization treatment of flue gas, and the desulfurization effect of the flue gas is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of flue gas desulfurization control, and in particular to a flue gas desulfurization control system and method. Background Art

[0002] Among the flue gas desulfurization technologies, wet desulfurization technology is relatively mature, with high efficiency and simple operation. This technology uses alkaline slurry (such as limestone, ammonia water) as the absorbent, which contacts the sulfur-containing flue gas in the absorption tower in countercurrent, and converts SO2 into sulfate through gas-liquid reaction.

[0003] Existing flue gas desulfurization control systems, such as a flue gas desulfurization control system for a thermal power unit disclosed in Chinese Patent Publication No. CN119847042A, include a thermal power unit data monitoring module, a thermal power unit data collection module, a thermal power unit data analysis module, a thermal power unit data processing module, and a thermal power unit data management module. The thermal power unit data monitoring module is responsible for real-time monitoring of flue gas composition, pipeline temperature, pipeline pressure, pollution content, and fault status during operation of the thermal power unit. However, the module cannot adaptively adjust the coverage of the mist liquid in the desulfurization tower, resulting in poor absorption of sulfur dioxide in the flue gas. Summary of the Invention

[0004] In order to overcome the above technical problems, the purpose of the present invention is to provide a flue gas desulfurization control system and method to solve the problem in the prior art that adaptive adjustment cannot be made according to the coverage rate of the mist liquid in the desulfurization tower, resulting in poor absorption effect of sulfur dioxide in the flue gas.

[0005] The purpose of the present invention can be achieved through the following technical solutions:

[0006] Specifically, the first aspect is to provide a flue gas desulfurization control system, including:

[0007] Pipeline module, which is used to collect flue gas data and slurry data;

[0008] A spray tower module, which is equipped with an ultrasonic generator and is used to collect slurry coverage;

[0009] a processing module configured to collect a historical training data set in advance and train a machine learning model for predicting slurry coverage in the spray tower module based on the historical training data set;

[0010] The pipeline module collects flue gas data and slurry data in real time. The processing module uses a machine learning module to predict the slurry coverage rate in the spray tower module based on the real-time collected flue gas data and slurry data, and adjusts the output power of the ultrasonic generator based on the predicted slurry coverage rate.

[0011] 2. A flue gas desulfurization control system according to claim 1, characterized in that the pipeline module includes a plurality of adjustable spray valves and spray pipelines, and the plurality of adjustable spray valves are distributed in multiple layers in the spray tower.

[0012] 3. A flue gas desulfurization control system according to claim 2, characterized in that a pressure sensor and a solenoid valve core are provided inside the adjustable spray valve.

[0013] 4. A flue gas desulfurization control system according to claim 1, characterized in that the flue gas data includes flue gas flow value, flue gas sulfur content percentage, flue gas temperature value and flue gas oxygen content percentage, and the slurry data includes slurry pressure value, slurry pH value and limestone particle size.

[0014] 5. A flue gas desulfurization control system according to claim 2, characterized in that the ultrasonic generators are distributed in multiple layers in the spray tower and correspond one-to-one to the multiple layers of the adjustable spray valve, and the ultrasonic generators are used to collect the slurry coverage inside the spray tower and perform ultrasonic treatment on the slurry inside the spray tower.

[0015] 6. A flue gas desulfurization control system according to claim 5, characterized in that the historical training data set includes N groups of training data, each group of training data includes feature data and label data, wherein the feature data includes flue gas flow value, flue gas sulfur percentage, flue gas temperature value, flue gas oxygen percentage, slurry pH value and limestone particle size, and the label data is slurry coverage.

[0016] 7. The flue gas desulfurization control system according to claim 6, wherein the method for the processing module to train a machine learning model for predicting the slurry coverage rate in the spray tower module comprises:

[0017] The feature data in the historical data is converted into feature vectors, and the feature vectors are used as the input of the machine learning model. The slurry coverage predicted by the machine learning model for the feature data is used as the output. The slurry coverage in the label data corresponding to each set of feature data is used as the prediction target. The machine learning model is trained until the model reaches convergence and training is stopped.

[0018] 8. A flue gas desulfurization control system according to claim 7, characterized in that the processing module generates a slurry diffusion change model based on the output power of the ultrasonic generator and the slurry coverage.

[0019] 9. A flue gas desulfurization control system according to claim 8, characterized in that the processing module uses a machine learning module to predict the slurry coverage in the spray tower module based on the real-time collected flue gas data, slurry data and slurry coverage, generates slurry diffusion based on the difference between the predicted slurry coverage and the preset slurry coverage, and the processing module uses a slurry diffusion change model based on the slurry diffusion to generate the output power of the ultrasonic generator.

[0020] A second specific aspect is to provide a flue gas desulfurization control method, which is implemented based on the above-mentioned flue gas desulfurization control system and includes the following steps:

[0021] S1: Pre-collect a historical training data set and train a machine learning model for predicting the slurry coverage rate in the spray tower module based on the historical training data set;

[0022] S2: Collect the slurry coverage and generate a slurry diffusion change model based on the output power of the ultrasonic generator and the slurry coverage;

[0023] S3: Collect flue gas data and slurry data in real time, and use a machine learning module to predict the slurry coverage in the spray tower module based on the real-time collected flue gas data and slurry data;

[0024] S4: Preset the slurry coverage of the spray tower, and generate the slurry diffusion based on the difference between the predicted slurry coverage and the preset slurry coverage;

[0025] S5: Generate the output power of the ultrasonic generator based on the slurry diffusivity using the slurry diffusivity change model;

[0026] S6: Determine the output power of the ultrasonic generator and perform ultrasonic treatment on the slurry inside the spray tower.

[0027] Beneficial effects of the present invention:

[0028] In the present invention, when the pipeline module collects flue gas data and slurry data in real time, the trained machine learning model is used to predict the slurry coverage rate in the spray tower module. When there is a gap between the slurry coverage rate and the preset slurry coverage rate, the output power of the ultrasonic generator is adjusted, and the ultrasonic generator is used to ultrasonically treat the mist liquid formed by the slurry in the internal space of the spray tower, and the distribution of the mist liquid is adjusted so that the slurry coverage rate is consistent with the desulfurization treatment of the flue gas, thereby improving the desulfurization effect of the flue gas. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] The present invention will be further described below with reference to the accompanying drawings.

[0030] Figure 1 This is a module block diagram of a flue gas desulfurization control system of the present invention;

[0031] Figure 2 This is a flow chart of a flue gas desulfurization control method of the present invention;

[0032] Figure 3 It is a schematic diagram of the internal structure of a spray tower in a flue gas desulfurization control system of the present invention.

[0033] Explanation of the accompanying drawings: 1. Spray tower; 2. Adjustable spray valve. DETAILED DESCRIPTION

[0034] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0035] As an embodiment of the present invention, Figure 1-Figure 3 As shown, the present invention discloses a flue gas desulfurization control system, including a pipeline module, a spray tower module and a processing module;

[0036] It should be noted that the spray pipe can be arranged in the middle position of the inner cavity of the spray tower 3. After passing through the side wall of the spray tower 3, the spray pipe is connected to the box for storing slurry in the external environment through a pipe, and a pump body for driving the flow of the slurry is provided in the pipe. The specifications and type of the pump body are adaptively selected by those skilled in the art according to the space size of the spray tower 3 and the spray pipe, so as to ensure that the slurry in the box can be stably supplied to the spray pipe in the middle position of the inner cavity of the spray tower 3 through the pipe. The installation position and the number of the pump body can be set arbitrarily, and the output power of the pump body can also be adaptively adjusted according to the state of the slurry in the spray pipe. The state of the slurry includes but is not limited to flow rate, temperature, viscosity, and composition.

[0037] In particular, if Figure 3 As shown, several adjustable spray valves 2 are divided into multiple groups, and each group is adaptively distributed according to the internal space position of the spray tower 3, so that the adjustable spray valves 2 can form a multi-layer state in the internal space of the spray tower 3. The adjustable spray valves in each layer are adaptively distributed according to the cross-section of the spray tower 3. The spacing of the adjustable spray valves 2 between the multiple layers can be fixed or adjustable. When adjustable, the spray pipe can be set to be retractable, that is, the length of the spray pipe can be adjusted. The retractable manner of the spray pipe can be adopted by those skilled in the art using retractable pipes in the prior art to ensure that the material of the pipe can provide the pressure of the slurry inside it;

[0038] The spacing between the adjustable spray valves 2 between the multiple layers can be adjusted adaptively according to the flue gas data entering the spray tower 3. For example, when the flue gas flow rate entering the spray tower 3 increases, the spacing between the adjustable spray valves between the multiple layers can be appropriately increased, that is, the contact space between the flue gas and the slurry is increased, thereby improving the adsorption effect of the slurry on sulfur dioxide in the flue gas. The specific adjustment range can be adaptively adjusted according to the actual situation of the spray tower 3.

[0039] It should also be emphasized that among the multiple adjustable spray valves 2 on one layer inside the spray tower 3, Figure 3 As shown, the adjustable spray valve 2 means that the position of the adjustable spray valve 2 in the cross section of the spray tower 3 (when the spray tower 3 is in a vertical state, that is, the horizontal plane) is adjustable. Specifically, each adjustable spray valve 2 can be connected to the spray pipe through a flexible pipe, and then the adjustable spray valve 2 is installed on the side of the spray tower 3 through a retractable structure. The retractable structure can be an electric push rod or a hydraulic push rod. The spatial position of the adjustable spray valve 2 in the horizontal plane is adjusted by the retractable structure.

[0040] A rectangular coordinate system is established with the center point of the cross section of the spray tower 3 (which is a horizontal plane when the spray tower 3 is in a vertical state) as the coordinate origin. The coordinate information of each adjustable spray valve 2 on the rectangular coordinate system and the coordinate interval information corresponding to the adjustable position of each adjustable spray valve 2 on the rectangular coordinate system are then marked.

[0041] For example, the coordinate information of the i-th adjustable spray valve 2 is (x i ,y i ), the corresponding coordinate interval information is:

[0042] y i =k i x i +b i , x i ∈[a i , c i ], y i ∈[d i , e i ];

[0043] where k i 、b i 、a i 、c i d i and e i are all constants, and their sizes are adaptively determined according to the specific position of the adjustable spray valve 2 on the rectangular coordinate system and the adjustable range;

[0044] Based on the above method, the coordinate information and coordinate interval information corresponding to all the adjustable spray valves 2 on the cross section of the spray tower 3 can be marked.

[0045] A pressure sensor and a solenoid valve core are installed inside the adjustable spray valve 2. The pressure sensor can be arbitrarily installed in the internal installation position of the adjustable spray valve 2, and its number can also be arbitrarily set. When multiple pressure sensors are installed inside the adjustable spray valve 2, the data used by the adjustable spray valve 2 can be the average value of the multiple pressure sensors. The specifications of the pressure sensor are adaptively selected according to the internal space of the adjustable spray valve 2. The pressure sensor installed inside the adjustable spray valve 2 can collect the slurry pressure value inside the valve 2.

[0046] The flue gas data includes the flue gas flow value, the flue gas sulfur percentage, the flue gas temperature value and the flue gas oxygen percentage. It should be noted that a flow meter and a thermometer can be set in the flue gas input pipe in the spray tower 3. The flow meter is used to collect the flue gas flow value entering the internal space of the spray tower 3, and the thermometer is used to collect the flue gas temperature value entering the internal space of the spray tower 3. The flue gas sulfur percentage and the flue gas oxygen percentage can be pre-measured or sampled in the flue gas input pipe, or can be collected in real time by sensors. The specific settings are arbitrarily made by those skilled in the art according to the actual on-site conditions of the spray tower 3. The flue gas sulfur percentage refers to the mass percentage of sulfur dioxide in the flue gas, and the flue gas oxygen percentage refers to the mass percentage of oxygen in the flue gas. Then, the flue gas flow value, flue gas sulfur percentage, flue gas temperature value and flue gas oxygen percentage can be converted into electrical signals and transmitted to the pipeline module;

[0047] An ultrasonic generator is provided inside the spray tower module, which is used to collect the slurry coverage. The ultrasonic generator is used to apply 20Hz to 100Hz ultrasonic waves to the internal space of the spray tower 3;

[0048] The slurry data includes a slurry pressure value, a slurry pH value, and a limestone particle size. The slurry pressure value refers to the slurry pressure value inside the adjustable spray valve 2, the slurry pH value refers to the slurry pH value of the tank storing the slurry, and the limestone particle size refers to the limestone particle size of the slurry in the tank storing the slurry. Both the slurry pH value and the limestone particle size can be adaptively pre-set according to the spray tower 3 and the flue gas data to ensure that sulfur dioxide in the flue gas can be adsorbed.

[0049] The processing module is used to collect a historical training data set in advance and train a machine learning model for predicting the slurry coverage rate in the spray tower module based on the historical training data set;

[0050] It should be noted that the slurry coverage of the mist formed by the slurry in the internal space of the spray tower 3 can be predicted by the machine learning model.

[0051] The pipeline module collects flue gas data and slurry data in real time, and the spray tower module collects slurry coverage in real time. The processing module uses the machine learning module to predict the slurry coverage in the spray tower module based on the real-time collected flue gas data, slurry data and slurry coverage, and adjusts the output power of the ultrasonic generator based on the slurry coverage. That is, when the pipeline module collects flue gas data and slurry data in real time, the trained machine learning model is used to predict the slurry coverage in the spray tower module. When there is a gap between the slurry coverage and the preset slurry coverage, the output power of the ultrasonic generator is adjusted, and the mist formed by the slurry in the internal space of the spray tower 3 is ultrasonically treated by the ultrasonic generator to adjust the distribution of the mist so that the slurry coverage is consistent with the desulfurization treatment of the flue gas, thereby improving the desulfurization effect of the flue gas.

[0052] As an embodiment of the present invention, Figure 1-Figure 3 As shown, the ultrasonic generators are distributed in multiple layers in the spray tower 3, and correspond one to one with the multiple layers of the adjustable spray valve 2. The ultrasonic generators are used to collect the slurry coverage inside the spray tower 3 and perform ultrasonic treatment on the slurry inside the spray tower 3. It should be noted that since the adjustable spray valve 2 is arranged in multiple layers in the internal space of the spray tower 3, the ultrasonic generator can fit the setting mode of the adjustable spray valve 2, and the ultrasonic generators are arranged in an array inside the spray tower 3. The slurry coverage of different layers inside the spray tower 3 is detected according to the difference in the transmission of ultrasonic waves in the mist liquid and the flue gas.

[0053] As an embodiment of the present invention, Figure 1-Figure 3 As shown, the historical training data set includes N groups of training data, where N is a positive integer, usually ranging from hundreds of thousands to millions. Each group of training data includes feature data and label data, wherein the feature data includes flue gas flow value, flue gas sulfur percentage, flue gas temperature value, flue gas oxygen percentage, slurry pH value and limestone particle size, and the label data is slurry coverage;

[0054] The flue gas flow rate value is the amount of flue gas entering the spray tower 3 when collecting each set of training data;

[0055] The sulfur content percentage of the flue gas is the mass percentage of sulfur dioxide in the flue gas entering the spray tower 3 when collecting each set of training data;

[0056] The flue gas temperature value is the flue gas temperature entering the spray tower 3 when collecting each set of training data;

[0057] The pH value of the slurry is the pH value of the slurry in the box when each set of training data is collected;

[0058] The limestone particle size is the limestone particle size of the slurry in the box when each set of training data was collected;

[0059] The label data is the slurry coverage rate of the mist liquid inside the spray tower 3 collected by the ultrasonic generator when collecting each set of training data.

[0060] The processing module trains a machine learning model to predict the slurry coverage in the spray tower module as follows:

[0061] The feature data in the historical data is converted into feature vectors, which are used as the input of the machine learning model. The machine learning model uses the slurry coverage predicted by the feature data as the output. The slurry coverage in the label data corresponding to each set of feature data is used as the prediction target. The training goal is to minimize the sum of the prediction accuracy. The calculation formula for prediction accuracy is ni = (si - zi) 2 , where ni is the prediction accuracy, si is the predicted slurry coverage corresponding to the i-th group of feature data, and zi is the slurry coverage in the label data corresponding to the i-th group of feature data. The machine learning model is trained until the model reaches convergence. The convergence standard is adaptively selected by those skilled in the art according to the specific model training situation.

[0062] As an embodiment of the present invention, Figure 1-Figure 3 As shown, the processing module generates a slurry diffusion change model based on the output power of the ultrasonic generator and the slurry coverage, specifically:

[0063] When a certain spray layer in the spray tower 3 switches and a transient pressure surge causes slurry backflow, it can be prevented by an ultrasonic generator. The relationship between the slurry coverage η and the ultrasonic frequency f of the ultrasonic generator is:

[0064]

[0065] Wherein, L is the vertical distance from the adjustable spray valve 2 to the inner wall of the spray tower 3, θ is the spray cone angle of the adjustable spray valve 2, D is the diameter of the spray tower 3, Q is the slurry flow rate value in a single adjustable spray valve 2, H is the height of a single spray layer in the spray tower 3, k is the empirical constant of droplet diameter (default k = 0.34), σ is the surface tension of the slurry, which is pre-determined by those skilled in the art, ρ is the density of the slurry, which is pre-determined by those skilled in the art, f is the ultrasonic frequency of a single ultrasonic generator, v gas is the flow rate of flue gas, which is measured according to the flue gas flow rate value;

[0066] in It represents the coverage ratio of the spray in the tower cross section, which is determined by the spray cone angle θ and the atomizer position L;

[0067] Indicates whether the slurry mist is sufficient to cover the target area. If the calculated result is ≥1, it means that the droplet density is saturated and the coverage rate is determined only by the area. If the calculated result is <1, the coverage rate needs to be corrected according to the actual density ratio;

[0068] The slurry coverage rate η can be directly calculated through the above formula. The slurry coverage rate η can also be changed by changing the size of the ultrasonic frequency f, that is, the slurry diffusion changes. When the slurry coverage rate η of a certain spray layer in the spray tower 3 cannot completely adsorb sulfur dioxide in the flue gas, the ultrasonic frequency f can be controlled according to the slurry diffusion change model to adjust the slurry coverage rate η of the spray layer in the spray tower 3.

[0069] As an embodiment of the present invention, Figure 1-Figure 3 As shown, the processing module uses a machine learning module to predict the slurry coverage in the spray tower module based on the real-time collected flue gas data, slurry data and slurry coverage, generates slurry diffusion based on the difference between the predicted slurry coverage and the preset slurry coverage, and generates the output power of the ultrasonic generator, that is, the ultrasonic frequency of the ultrasonic generator, based on the slurry diffusion using a slurry diffusion change model;

[0070] In particular, the position of each adjustable spray valve 2 in the spray layer of the spray tower 3 can be controlled according to the slurry diffusion degree generated based on the difference between the predicted slurry coverage and the preset slurry coverage;

[0071] Specifically, according to the coordinate information and coordinate interval information corresponding to each adjustable spray valve 2 in the marked spray layer, a slurry diffusion change model is generated according to the change of the coordinate interval information of each adjustable spray valve 2. When there is a difference between the predicted slurry coverage rate and the preset slurry coverage rate, the position of each adjustable spray valve 2 can be adjusted by controlling the coordinate information corresponding to the coordinate interval information of the adjustable spray valve 2, so as to achieve the purpose of controlling the slurry coverage rate, ensure that the sulfur dioxide in the flue gas in the spray tower 3 can be completely adsorbed, and improve the adsorption effect of sulfur dioxide.

[0072] It should also be emphasized that the spray layers in the spray tower 3 are set in multiple layers. Since the flue gas in the spray tower 3 flows from bottom to top, and the mist formed by the slurry flows from top to bottom in the spray tower 3, taking the five spray layers in the spray tower 3 as an example, the five spray layers are marked as 1-5 from bottom to top. When the slurry coverage rate generated by the adjustable spray valve 2 in the spray layer No. 1 deviates, the adjustment of the spray layer No. 1 by the ultrasonic generator corresponding to the spray layer No. 1 has a hysteresis, so the ultrasonic generator corresponding to the spray layer No. 2 can be adjusted accordingly. For example, according to the difference between the predicted slurry coverage rate of the spray layer No. 1 and the preset slurry coverage rate, the ultrasonic generator corresponding to the spray layer No. 2 is controlled to increase the corresponding ultrasonic frequency. In this way, the defects generated in the spray layer No. 1 can be supplemented in time to ensure the adsorption effect of sulfur dioxide in the flue gas.

[0073] As an embodiment of the present invention, Figure 1-Figure 3 As shown, a flue gas desulfurization control method is implemented based on the above-mentioned flue gas desulfurization control system, comprising the following steps:

[0074] S1: Pre-collect a historical training data set, and train a machine learning model for predicting the slurry coverage rate in the spray tower module based on the historical training data set. Specifically, the historical training data set includes N groups of training data, where N is a positive integer, typically ranging from hundreds of thousands to millions, and each group of training data includes feature data and label data, wherein the feature data includes flue gas flow rate value, flue gas sulfur content percentage, flue gas temperature value, flue gas oxygen content percentage, slurry pH value, and limestone particle size, and the label data is the slurry coverage rate;

[0075] The flue gas flow rate value is the amount of flue gas entering the spray tower 3 when collecting each set of training data;

[0076] The sulfur content percentage of the flue gas is the mass percentage of sulfur dioxide in the flue gas entering the spray tower 3 when collecting each set of training data;

[0077] The flue gas temperature value is the flue gas temperature entering the spray tower 3 when collecting each set of training data;

[0078] The pH value of the slurry is the pH value of the slurry in the box when each set of training data is collected;

[0079] The limestone particle size is the limestone particle size of the slurry in the box when each set of training data was collected;

[0080] The label data is the slurry coverage rate of the mist liquid inside the spray tower 3 collected by the ultrasonic generator when collecting each set of training data.

[0081] The processing module trains a machine learning model to predict the slurry coverage in the spray tower module as follows:

[0082] The feature data in the historical data is converted into feature vectors, which are used as the input of the machine learning model. The machine learning model uses the slurry coverage predicted by the feature data as the output. The slurry coverage in the label data corresponding to each set of feature data is used as the prediction target. The training goal is to minimize the sum of the prediction accuracy. The calculation formula for prediction accuracy is ni = (si - zi) 2 , where ni is the prediction accuracy, si is the predicted slurry coverage corresponding to the i-th set of feature data, and zi is the slurry coverage in the label data corresponding to the i-th set of feature data. The machine learning model is trained until the model reaches convergence. The convergence standard is adaptively selected by those skilled in the art according to the specific model training situation.

[0083] S2: Collect the slurry coverage rate and generate a slurry diffusion change model based on the output power of the ultrasonic generator and the slurry coverage rate. Specifically:

[0084] When a certain spray layer in the spray tower 3 switches and a transient pressure surge causes slurry backflow, it can be prevented by an ultrasonic generator. The relationship between the slurry coverage η and the ultrasonic frequency f of the ultrasonic generator is:

[0085]

[0086] Wherein, L is the vertical distance from the adjustable spray valve 2 to the inner wall of the spray tower 3, θ is the spray cone angle of the adjustable spray valve 2, D is the diameter of the spray tower 3, Q is the slurry flow rate value in a single adjustable spray valve 2, H is the height of a single spray layer in the spray tower 3, k is the empirical constant of droplet diameter (default k = 0.34), σ is the surface tension of the slurry, which is pre-determined by those skilled in the art, ρ is the density of the slurry, which is pre-determined by those skilled in the art, f is the ultrasonic frequency of a single ultrasonic generator, v gas is the flow rate of flue gas, which is measured according to the flue gas flow rate value;

[0087] in It represents the coverage ratio of the spray in the tower cross section, which is determined by the spray cone angle θ and the atomizer position L;

[0088] Indicates whether the slurry mist is sufficient to cover the target area. If the calculated result is ≥1, it means that the droplet density is saturated and the coverage rate is determined only by the area. If the calculated result is <1, the coverage rate needs to be corrected according to the actual density ratio;

[0089] The slurry coverage rate η can be directly calculated using the above formula. The slurry coverage rate η can also be changed by changing the ultrasonic frequency f, that is, the slurry diffusion rate changes. When the slurry coverage rate η of a certain spray layer in the spray tower 3 cannot completely adsorb sulfur dioxide in the flue gas, the ultrasonic frequency f can be controlled according to the slurry diffusion rate change model to adjust the slurry coverage rate η of the spray layer in the spray tower 3.

[0090] S3: real-time collection of flue gas data and slurry data, and prediction of the slurry coverage in the spray tower module using a machine learning module based on the real-time collected flue gas data and slurry data. Specifically, the spray pipe can be set in the middle of the inner cavity of the spray tower 3. The spray pipe passes through the side wall of the spray tower 3 and is connected to the box for storing slurry in the external environment through a pipe, and a pump body for driving the flow of slurry is set in the pipe. The specifications and type of the pump body are adaptively selected by those skilled in the art according to the space size of the spray tower 3 and the spray pipe to ensure that the slurry in the box can be stably supplied to the spray pipe in the middle of the inner cavity of the spray tower 3 through the pipe. The installation position and number of the pump body can be set arbitrarily, and the output power of the pump body can also be adaptively adjusted according to the state of the slurry in the spray pipe. The state of the slurry includes but is not limited to flow rate, temperature, viscosity, and composition.

[0091] In particular, if Figure 3 As shown, several adjustable spray valves 2 are divided into multiple groups, and each group is adaptively distributed according to the internal space position of the spray tower 3, so that the adjustable spray valves 2 can form a multi-layer state in the internal space of the spray tower 3. The adjustable spray valves in each layer are adaptively distributed according to the cross-section of the spray tower 3. The spacing of the adjustable spray valves 2 between the multiple layers can be fixed or adjustable. When adjustable, the spray pipe can be set to be retractable, that is, the length of the spray pipe can be adjusted. The retractable manner of the spray pipe can be adopted by those skilled in the art using retractable pipes in the prior art to ensure that the material of the pipe can provide the pressure of the slurry inside it;

[0092] The spacing between the adjustable spray valves 2 between the multiple layers can be adjusted adaptively according to the flue gas data entering the spray tower 3. For example, when the flue gas flow rate entering the spray tower 3 increases, the spacing between the adjustable spray valves between the multiple layers can be appropriately increased, that is, the contact space between the flue gas and the slurry is increased, thereby improving the adsorption effect of the slurry on sulfur dioxide in the flue gas. The specific adjustment range can be adaptively adjusted according to the actual situation of the spray tower 3.

[0093] It should also be emphasized that among the multiple adjustable spray valves 2 on one layer inside the spray tower 3, Figure 3As shown, the adjustable spray valve 2 means that the position of the adjustable spray valve 2 in the cross section of the spray tower 3 (when the spray tower 3 is in a vertical state, that is, the horizontal plane) is adjustable. Specifically, each adjustable spray valve 2 can be connected to the spray pipe through a flexible pipe, and then the adjustable spray valve 2 is installed on the side of the spray tower 3 through a retractable structure. The retractable structure can be an electric push rod or a hydraulic push rod. The spatial position of the adjustable spray valve 2 in the horizontal plane is adjusted by the retractable structure.

[0094] A rectangular coordinate system is established with the center point of the cross section of the spray tower 3 (which is a horizontal plane when the spray tower 3 is in a vertical state) as the coordinate origin. The coordinate information of each adjustable spray valve 2 on the rectangular coordinate system and the coordinate interval information corresponding to the adjustable position of each adjustable spray valve 2 on the rectangular coordinate system are then marked.

[0095] For example, the coordinate information of the i-th adjustable spray valve 2 is (x i ,y i ), the corresponding coordinate interval information is:

[0096] y i =k i x i +b i , x i ∈[a i , c i ], y i ∈[d i , e i ];

[0097] where k i 、b i 、a i 、c i d i and e i are all constants, and their sizes are adaptively determined according to the specific position of the adjustable spray valve 2 on the rectangular coordinate system and the adjustable range;

[0098] Based on the above method, the coordinate information and coordinate interval information corresponding to all the adjustable spray valves 2 on the cross section of the spray tower 3 can be marked.

[0099] A pressure sensor and a solenoid valve core are installed inside the adjustable spray valve 2. The pressure sensor can be arbitrarily installed in the internal installation position of the adjustable spray valve 2, and its number can also be arbitrarily set. When multiple pressure sensors are installed inside the adjustable spray valve 2, the data used by the adjustable spray valve 2 can be the average value of the multiple pressure sensors. The specifications of the pressure sensor are adaptively selected according to the internal space of the adjustable spray valve 2. The pressure sensor installed inside the adjustable spray valve 2 can collect the slurry pressure value inside the valve 2.

[0100] The flue gas data includes the flue gas flow value, the flue gas sulfur percentage, the flue gas temperature value and the flue gas oxygen percentage. It should be noted that a flow meter and a thermometer can be set in the flue gas input pipe in the spray tower 3. The flow meter is used to collect the flue gas flow value entering the internal space of the spray tower 3, and the thermometer is used to collect the flue gas temperature value entering the internal space of the spray tower 3. The flue gas sulfur percentage and the flue gas oxygen percentage can be pre-measured or sampled in the flue gas input pipe, or can be collected in real time by sensors. The specific settings are arbitrarily made by those skilled in the art according to the actual on-site conditions of the spray tower 3. The flue gas sulfur percentage refers to the mass percentage of sulfur dioxide in the flue gas, and the flue gas oxygen percentage refers to the mass percentage of oxygen in the flue gas. Then, the flue gas flow value, flue gas sulfur percentage, flue gas temperature value and flue gas oxygen percentage can be converted into electrical signals and transmitted to the pipeline module;

[0101] An ultrasonic generator is provided inside the spray tower module, which is used to collect the slurry coverage. The ultrasonic generator is used to apply 20Hz to 100Hz ultrasonic waves to the internal space of the spray tower 3;

[0102] The slurry data includes a slurry pressure value, a slurry pH value, and a limestone particle size. The slurry pressure value refers to the slurry pressure value inside the adjustable spray valve 2, the slurry pH value refers to the slurry pH value of the tank storing the slurry, and the limestone particle size refers to the limestone particle size of the slurry in the tank storing the slurry. Both the slurry pH value and the limestone particle size can be adaptively pre-set according to the spray tower 3 and the flue gas data to ensure that sulfur dioxide in the flue gas can be adsorbed.

[0103] S4: Preset the slurry coverage of the spray tower, and generate the slurry diffusion based on the difference between the predicted slurry coverage and the preset slurry coverage;

[0104] S5: Based on the slurry diffusivity, the output power of the ultrasonic generator is generated using a slurry diffusivity change model. The processing module uses a machine learning module to predict the slurry coverage in the spray tower module based on the real-time collected flue gas data, slurry data, and slurry coverage. The slurry diffusivity is generated based on the difference between the predicted slurry coverage and the preset slurry coverage. The processing module uses the slurry diffusivity change model to generate the output power of the ultrasonic generator based on the slurry diffusivity, that is, the ultrasonic frequency of the ultrasonic generator;

[0105] S6: Determine the output power of the ultrasonic generator and perform ultrasonic treatment on the slurry inside the spray tower 3 so that the slurry coverage inside the spray tower 3 meets the preset value, thereby ensuring the adsorption effect of the slurry on sulfur dioxide in the flue gas.

[0106] The above is a detailed description of an embodiment of the present invention. However, the content described is only a preferred embodiment of the present invention and should not be considered to limit the scope of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A flue gas desulfurization control system, characterized in that: include: Pipeline module, which is used to collect flue gas data and slurry data; A spray tower module, which is equipped with an ultrasonic generator and is used to collect slurry coverage; a processing module configured to collect a historical training data set in advance and train a machine learning model for predicting slurry coverage in the spray tower module based on the historical training data set; The pipeline module collects flue gas data and slurry data in real time. The processing module uses a machine learning module to predict the slurry coverage rate in the spray tower module based on the real-time collected flue gas data and slurry data, and adjusts the output power of the ultrasonic generator based on the predicted slurry coverage rate.

2. A flue gas desulfurization control system according to claim 1, characterized in that: The pipeline module includes a plurality of adjustable spray valves and spray pipelines, and the plurality of adjustable spray valves are distributed in multiple layers in the spray tower.

3. A flue gas desulfurization control system according to claim 2, characterized in that: A pressure sensor and a solenoid valve core are arranged inside the adjustable spray valve.

4. A flue gas desulfurization control system according to claim 1, characterized in that: The flue gas data includes flue gas flow rate value, flue gas sulfur content percentage, flue gas temperature value and flue gas oxygen content percentage, and the slurry data includes slurry pressure value, slurry pH value and limestone particle size.

5. A flue gas desulfurization control system according to claim 2, characterized in that: The ultrasonic generators are distributed in multiple layers in the spray tower and correspond one-to-one to the multiple layers of the adjustable spray valves. The ultrasonic generators are used to collect the slurry coverage rate inside the spray tower and perform ultrasonic treatment on the slurry inside the spray tower.

6. A flue gas desulfurization control system according to claim 5, characterized in that: The historical training data set includes N groups of training data, each group of training data includes feature data and label data, wherein the feature data includes flue gas flow value, flue gas sulfur percentage, flue gas temperature value, flue gas oxygen percentage, slurry pH value and limestone particle size, and the label data is slurry coverage.

7. A flue gas desulfurization control system according to claim 6, characterized in that: The method for the processing module to train a machine learning model for predicting slurry coverage in the spray tower module includes: The feature data in the historical data is converted into feature vectors, and the feature vectors are used as the input of the machine learning model. The slurry coverage predicted by the machine learning model for the feature data is used as the output. The slurry coverage in the label data corresponding to each set of feature data is used as the prediction target. The machine learning model is trained until the model reaches convergence and training is stopped.

8. A flue gas desulfurization control system according to claim 7, characterized in that: The processing module generates a slurry diffusion variation model based on the output power of the ultrasonic generator and the slurry coverage.

9. A flue gas desulfurization control system according to claim 8, characterized in that: The processing module uses a machine learning module to predict the slurry coverage in the spray tower module based on the real-time collected flue gas data, slurry data and slurry coverage, generates slurry diffusion based on the difference between the predicted slurry coverage and the preset slurry coverage, and generates the output power of the ultrasonic generator based on the slurry diffusion using a slurry diffusion change model.

10. A flue gas desulfurization control method, characterized in that: The method is implemented based on a flue gas desulfurization control system according to any one of claims 1 to 9, and comprises the following steps: S1: Collect a historical training data set in advance and train a machine learning model for predicting the slurry coverage rate in the spray tower module based on the historical training data set; S2: Collect slurry coverage and generate a slurry diffusion change model based on the output power of the ultrasonic generator and the slurry coverage; S3: Collect flue gas data and slurry data in real time, and use the machine learning module to predict the slurry coverage in the spray tower module based on the real-time collected flue gas data and slurry data; S4: Preset the slurry coverage of the spray tower, and generate the slurry diffusion based on the difference between the predicted slurry coverage and the preset slurry coverage; S5: Generate the output power of the ultrasonic generator based on the slurry diffusivity using the slurry diffusivity change model; S6: Determine the output power of the ultrasonic generator and perform ultrasonic treatment on the slurry inside the spray tower.

Images