Provides an apparatus, method, and computer program product for recommending the dosage of desulfurization to a flue gas desulfurization system.

Abstract

An apparatus, method, and computer program product for providing recommended desulfurization dosages to flue gas desulfurization (FGD) equipment. The FGD equipment includes a desulfurization slurry tank and a desulfurization reaction zone. The apparatus performs the following operations for each candidate desulfurization dosage: (a) inputting the candidate desulfurization dosage and the operating conditions of the FGD equipment into a pH prediction model of the desulfurization slurry tank to obtain a predicted pH value; (b) inputting the candidate desulfurization dosage, the combination of operating conditions, and the predicted pH value into a sulfur oxide concentration prediction model of the desulfurization reaction zone to obtain a predicted sulfur oxide concentration; and (c) calculating an evaluation value based on the difference between the predicted pH value and the target pH value, the difference between the predicted sulfur oxide concentration and the target sulfur oxide concentration, and the candidate desulfurization dosage. The apparatus further determines a recommended desulfurization dosage from the candidate desulfurization dosages based on these evaluation values.

Description

[Technical Field]

[0001] This invention relates to an apparatus, method, and computer program product for providing a recommended desulfurization dosage to a flue gas desulfurization (FGD) device. Specifically, this invention relates to an apparatus, method, and computer program product for providing a recommended desulfurization dosage to a FGD device based on a combination of operating conditions of the FGD device. [Previous Technology]

[0002] The main processes of a cogeneration plant (e.g., a coal-fired power plant) involve boilers, turbines, feedwater systems, cooling systems, and environmental protection equipment. The heat generated by burning coal in the boiler heats saturated water into hot steam, which is then directed to the turbine to generate electricity. Additionally, low-pressure steam can be extracted for use by other equipment. The flue gas produced after burning coal in the boiler contains pollutants such as nitrogen oxides (NOx), fly ash, and sulfur oxides (SOx). These pollutants are treated by environmental protection equipment for denitrification, dust removal, and desulfurization before being emitted into the atmosphere through the chimney.

[0003] To ensure that the concentration of sulfur oxides in the flue gas emitted by environmental protection equipment complies with environmental regulations, some cogeneration plants install flue gas desulfurization (FGD) equipment to remove sulfur from the flue gas. Figure 1 depicts a schematic diagram of a typical FGD equipment 1, which is a desulfurization tower. At the bottom of the FGD equipment 1 is a desulfurization slurry tank T1 for holding magnesium hydroxide slurry. The magnesium hydroxide slurry can be added to the desulfurization slurry tank T1 by operating the control valve P2. The circulating pump P1 guides the magnesium hydroxide slurry from the desulfurization slurry tank T1 into the FGD equipment 1. Sprayers s1, s2, and s3 then spray the magnesium hydroxide slurry onto the perforated plates B1 and B2 inside the FGD equipment 1 to form a liquid film. The sprayed magnesium hydroxide slurry reacts with the oxides in the flue gas S introduced into the FGD equipment 1, achieving the purpose of removing sulfur oxides.

[0004] However, using flue gas desulfurization equipment for flue gas desulfurization faces several challenges. First, the effectiveness of flue gas desulfurization is affected by many factors, such as boiler load changes, coal quality, slurry circulation volume, slurry pH, magnesium oxide purity, causticization rate, and blower volume. Therefore, the concentration of sulfur oxides at the outlet of the flue gas desulfurization equipment fluctuates greatly, requiring continuous adjustment of the dosage, which is difficult to control. Second, fresh magnesium hydroxide slurry added to the desulfurization slurry tank needs to undergo liquid-phase diffusion and circulation pump mixing, thus delaying the desulfurization reaction. Third, the flue gas desulfurization equipment adds magnesium hydroxide slurry to the desulfurization slurry tank through the opening and closing of control valves. Since magnesium hydroxide slurry is a suspended particulate liquid with abrasive properties, control valves are easily damaged and leak. Therefore, operators need to intermittently open and close the control valves to mitigate damage, making continuous proportional control impossible. This results in operators frequently adjusting the opening and closing of control valves to adjust the amount of magnesium hydroxide slurry added, increasing their workload. Fourth, to ensure that the emitted flue gas complies with environmental regulations, operators often increase the amount of magnesium hydroxide slurry used, leading to increased operating costs. Fifth, excessive addition of magnesium hydroxide slurry increases the suspended solids in the wastewater discharged from the flue gas desulfurization equipment, causing blockages in wastewater pipelines, which in turn increases the load on downstream sludge treatment and raises sludge removal costs.

[0005] In view of this, there is an urgent need in the art for a technology that can provide recommended desulfurization dosage for flue gas desulfurization equipment according to different operating conditions in order to solve the aforementioned problems. [Summary of the Invention]

[0006] One object of the present invention is to provide an apparatus for providing a suggested desulfurization dose to a flue gas desulfurization (FGD) device. The FGD device includes a desulfurization slurry tank and a desulfurization reaction zone. The apparatus includes a storage device and a processor, wherein the processor is electrically connected to the storage device. The storage device stores an pH prediction model of the desulfurization slurry tank and a sulfur oxide concentration prediction model of the desulfurization reaction zone. For each of a plurality of candidate desulfurization doses, the following operations are performed: (a) combining the candidate desulfurization dose with an operating condition of the FGD device and inputting it into the pH prediction model to obtain a predicted pH; (b) inputting the candidate desulfurization dose, the operating condition combination, and the predicted pH into the sulfur oxide concentration prediction model to obtain a predicted sulfur oxide concentration; and (c) calculating an evaluation value based on a first difference between the predicted pH and a target pH, a second difference between the predicted sulfur oxide concentration and a target sulfur oxide concentration, and the candidate desulfurization dose. The processor further determines the recommended desulfurization dose from among the candidate desulfurization doses based on these evaluation values.

[0007] Another object of the present invention is to provide a method for providing a recommended desulfurization dosage to a flue gas desulfurization device, the method being executed by an electronic computing device. The flue gas desulfurization device includes a desulfurization slurry tank and a desulfurization reaction zone. The electronic computing device stores a pH prediction model of the desulfurization slurry tank and a sulfur oxide concentration prediction model of the desulfurization reaction zone. The method comprises the following steps: (a) performing the following steps for each of a plurality of candidate desulfurization doses: (a1) inputting the candidate desulfurization dose and an operating condition of the flue gas desulfurization equipment into the pH prediction model to obtain a predicted pH; (a2) inputting the candidate desulfurization dose, the operating condition combination, and the predicted pH into the sulfur oxide concentration prediction model to obtain a predicted sulfur oxide concentration; and (a3) ​​calculating an evaluation value based on a first difference between the predicted pH and a target pH, a second difference between the predicted sulfur oxide concentration and a target sulfur oxide concentration, and the candidate desulfurization dose; and (b) determining the recommended desulfurization dose from the candidate desulfurization doses based on the evaluation value.

[0008] Another object of the present invention is to provide a computer program product. After the computer program product is loaded via an electronic computing device, the electronic computing device executes a plurality of program instructions contained in the computer program product to implement the method described above.

[0009] The present invention provides a technique (including at least an apparatus, method, and computer program product) for providing a suggested desulfurization dosage to a flue gas desulfurization (FGD) device. In the process of calculating the evaluation values ​​of each candidate desulfurization dosage, the method first uses an acid-base value prediction model to predict the acid-base value of the desulfurization slurry tank, and then uses a sulfur oxide concentration prediction model to predict the sulfur oxide concentration at the outlet of the FGD device based on the prediction results of the acid-base value prediction model. By employing the acid-base value prediction model and the acid-base value prediction model executed sequentially, the present invention has considered the situation where the acid-base value of the desulfurization slurry tank reacts faster after the addition of magnesium hydroxide slurry, while the sulfur oxide concentration at the outlet of the FGD device reacts slower. Furthermore, the present invention takes into account various achievable benefits when calculating the evaluation value of each candidate desulfurization dose, including: the degree of closeness or difference with the target pH value of the desulfurization slurry pool (reflected by the first difference), the degree of closeness or difference with the target sulfur oxide concentration at the flue gas desulfurization equipment outlet (reflected by the second difference), and the dosage of the added desulfurizing agent (reflected by the candidate desulfurization dose). Therefore, the present invention can solve all the problems of the prior art.

[0010] The following detailed description of the technology and implementation of the present invention is illustrated in conjunction with the drawings, so that those skilled in the art can understand the technical features of the claimed invention.

Implementation Method

[0011] The following will explain the apparatus, method, and computer program product for providing a suggested desulfurization dosage to a flue gas desulfurization device according to the present invention through embodiments. However, these embodiments are not intended to limit the implementation of the present invention to any environment, application, or manner described in these embodiments. The description of the following embodiments is only for illustrating the purpose of the present invention and is not intended to limit the scope of the present invention. It should be understood that in the following embodiments and drawings, elements not directly related to the technical features of the present invention have been omitted and not described or / and illustrated. In addition, the dimensions of the elements and the proportional relationships between the elements in the drawings are only for illustration and explanation and are not intended to limit the scope of the present invention. Furthermore, unless otherwise stated, the terms "a," "the," and similar terms used in this specification and the claims should be understood to include both singular and plural forms.

[0012] A recommended desulfurization dosage is provided for a flue gas desulfurization system.

[0013] Figure 2 depicts a schematic diagram of the architecture of a desulfurization dosage recommendation device (i.e., a device for providing a recommended desulfurization dosage to a flue gas desulfurization device) 2 according to some embodiments of the present invention. The desulfurization dosage recommendation device 2 includes a storage unit 21 and a processor 23, with the processor 23 electrically connected to the storage unit 21. The storage unit 21 may include one or more of the following: memory, hard disk, and any other non-transitory storage medium, circuitry, or device with the same function known to those skilled in the art to which this invention pertains. The processor 23 may include one or more of the following: various processors, central processing units (CPUs), microprocessors (MPUs), digital signal processors (DSPs), graphics processing units (GPUs), and other computing devices known to those skilled in the art to which this invention pertains. In some embodiments, the desulfurization dosage recommendation device 2 may also include an input interface 25, with the input interface 25 electrically connected to the processor 23. The input interface 25 may include one or more of the following: keyboard, mouse, touch screen, trackball, voice converter and other devices that allow the user to input commands.

[0014] The following will use the flue gas desulfurization equipment 1 shown in Figure 1 as an example to explain in detail the operation of the desulfurization dosage recommendation device 2. However, it should be understood that the present invention can also be applied to other flue gas desulfurization equipment. As mentioned above, the flue gas desulfurization equipment 1 includes a desulfurization slurry tank T1. In addition, the flue gas desulfurization equipment 1 has a desulfurization reaction zone A (i.e., the area where the magnesium hydroxide slurry sprayed by sprayers s1, s2, and s3 reacts with the oxides in the flue gas S introduced into the flue gas desulfurization equipment 1), the range of which generally covers the outlets of sprayers s1, s2, and s3 to the lowest perforated plate B2. In the desulfurization reaction zone A, the magnesium hydroxide slurry reacts with the flue gas S with the following chemical formulas (1), (2), and / or (3). (1)(2)(3)

[0015] The reaction performed by the flue gas desulfurization equipment 1 is an acid-base neutralization reaction. The sulfur oxide concentration at the outlet of the flue gas desulfurization equipment 1 is not only closely inversely related to the pH value of the desulfurization slurry tank T1, but is also affected by other parameters; therefore, the relationship between the two is not entirely linear. Furthermore, after adding magnesium hydroxide slurry to the desulfurization slurry tank T1, the pH value of the desulfurization slurry tank T1 changes relatively quickly, but the sulfur oxide concentration at the outlet of the flue gas desulfurization equipment 1 changes more slowly. To comprehensively consider the aforementioned factors, it is recommended that the desulfurization dosage device 2 adopt a trained and validated pH value prediction model M1 and a sulfur oxide concentration prediction model M2, and the prediction results of the pH value prediction model M1 will be used in the sulfur oxide concentration prediction model M2.

[0016] Specifically, the storage device 21 stores a pH prediction model M1 for the desulfurization slurry tank T1 and a sulfur oxide concentration prediction model M2 for the desulfurization reaction zone A. Specifically, the pH prediction model M1 is trained and validated to predict the pH of the desulfurization slurry tank T1 when the flue gas desulfurization equipment 1 operates with the desulfurization dosage (e.g., the dosage of magnesium hydroxide slurry) and the combination of operating conditions of the flue gas desulfurization equipment 1 (i.e., the values ​​of a plurality of given variables or characteristic variables corresponding to the flue gas desulfurization equipment 1). Furthermore, the sulfur oxide concentration prediction model M2, after training and validation, can predict the sulfur oxide concentration in the desulfurization reaction zone A (i.e., the sulfur oxide concentration at the chimney outlet of the flue gas desulfurization equipment 1) when the flue gas desulfurization equipment 1 operates at a given desulfurization dosage and operating condition combination based on the desulfurization dosage of the desulfurization slurry tank T1 and the combination of operating conditions of the flue gas desulfurization equipment 1. It should be noted that these given variables of the flue gas desulfurization equipment 1 are controllable or observable variables during the operation of the flue gas desulfurization equipment 1, while these characteristic variables of the flue gas desulfurization equipment 1 are the most critical among these given variables in terms of their impact on the flue gas desulfurization equipment 1 (explained later).

[0017] For the combination of operating conditions 10 during the operation of the flue gas desulfurization equipment 1, the desulfurization dosage recommendation device 2 will use the pH prediction model M1 and the sulfur oxide concentration prediction model M2 to evaluate and then provide a recommended desulfurization dosage (not shown). Please refer to Figure 3 for the following explanation, which depicts the main data flow for providing the recommended desulfurization dosage for the flue gas desulfurization equipment 1.

[0018] The combination of operating conditions 10 during the operation of the flue gas desulfurization equipment 1 is known. For example, the combination of operating conditions 10 may include the main steam flow rate of the flue gas desulfurization equipment 1, the absorption tower circulation rate, the sulfur oxide content at the chimney inlet, the flue gas flow rate at the chimney outlet, the opening time of control valve P2 (equivalent to the time of adding magnesium hydroxide slurry, which can also be regarded as the dosage of magnesium hydroxide slurry), the blower air volume, the magnesium hydroxide slurry discharge rate, the magnesium hydroxide slurry temperature, and / or the absorption tower makeup water volume. The desulfurization dosage recommendation device 2 obtains the combination of operating conditions 10 during the operation of the flue gas desulfurization equipment 1. It should be noted that the present invention does not limit how the desulfurization dosage recommendation device 2 obtains the combination of operating conditions 10 during the operation of the flue gas desulfurization equipment 1. In some embodiments, the user can input the combination of operating conditions 10 into the desulfurization dosage recommendation device 2 through the input interface 25. In other embodiments, the desulfurization dosage recommendation device 2 may be configured with a network interface and receive the combination of operating conditions 10 through the network interface.

[0019] The storage unit 21 stores a plurality of candidate desulfurization doses A_1, ..., A_N. For each of the candidate desulfurization doses A_1, ..., A_N, the processor 23 performs operation (a), operation (b) and operation (c).

[0020] In operation (a), processor 23 inputs the candidate desulfurization dose and the operating condition combination 10 of flue gas desulfurization equipment 1 into the pH prediction model M1 to obtain the predicted pH value. As shown in Figure 3, if the current evaluation is of candidate desulfurization dose A_1, processor 23 inputs candidate desulfurization dose A_1 and the operating condition combination 10 into the pH prediction model M1 to obtain the predicted pH value F1_1. If the current evaluation is of candidate desulfurization dose A_N, processor 23 inputs candidate desulfurization dose A_N and the operating condition combination 10 into the pH prediction model M1 to obtain the predicted pH value F1_N. For other candidate desulfurization doses, processor 23 will also perform similar operations to obtain the corresponding predicted pH value.

[0021] In operation (b), processor 23 inputs the candidate desulfurization dose, operating condition combination 10, and the predicted pH value into sulfur oxide concentration prediction model M2 to obtain the predicted sulfur oxide concentration. As shown in Figure 3, if the current evaluation is of candidate desulfurization dose A_1, processor 23 inputs candidate desulfurization dose A_1, operating condition combination 10, and predicted pH value F1_1 into sulfur oxide concentration prediction model M2 to obtain the predicted sulfur oxide concentration F2_1. If the current evaluation is of candidate desulfurization dose A_N, processor 23 inputs candidate desulfurization dose A_N, operating condition combination 10, and predicted pH value F1_N into sulfur oxide concentration prediction model M2 to obtain the predicted sulfur oxide concentration F2_N. For other candidate desulfurization doses, processor 23 will also perform similar operations to obtain the corresponding predicted sulfur oxide concentration.

[0022] In operation (c), processor 23 calculates an evaluation value based on a first difference between the predicted pH and a target pH O1, a second difference between the predicted sulfur oxide concentration and a target sulfur oxide concentration O2, and the candidate desulfurization dose. The target pH O1 is a value that allows the flue gas desulfurization equipment 1 to operate safely (e.g., between 6 and 8), and the target sulfur oxide concentration O2 is a value that complies with environmental regulations. As shown in Figure 3, if the candidate desulfurization dose A_1 is currently being evaluated, processor 23 calculates the OP evaluation value E_1 based on the first difference D1_1 between the predicted pH F1_1 and the target pH O1, the second difference D2_1 between the predicted sulfur oxide concentration F2_1 and the target sulfur oxide concentration O2, and the candidate desulfurization dose A_1. If the current evaluation is of candidate desulfurization dose A_N, processor 23 calculates the OP evaluation value E_N based on the first difference D1_N between the predicted pH value F1_N and the target pH value O1, the second difference D2_N between the predicted sulfur oxide concentration F2_N and the target sulfur oxide concentration O2, and the candidate desulfurization dose A_N. For other candidate desulfurization doses, processor 23 will perform similar operations to obtain the corresponding evaluation values.

[0023] It should be noted that the first difference corresponding to a candidate desulfurization dose can be used to reflect the degree of closeness (or difference) between the predicted pH value and the target pH value O1. Furthermore, the second difference corresponding to a candidate desulfurization dose can be used to reflect the degree of closeness (or difference) between the predicted sulfur oxide concentration and the target sulfur oxide concentration O2. Moreover, the magnitude of a candidate desulfurization dose can be used to reflect whether the control valve P2 is easily cut and leaks, and the amount of suspended solids in the wastewater discharged by the flue gas desulfurization equipment. Therefore, in different embodiments, the processor 23 can use different objective functions to calculate the evaluation value based on the first difference, the second difference, and the candidate desulfurization dose.

[0024] In some embodiments, the evaluation value calculated by the processor 23 for a candidate desulfurization dose is the sum of the square of the first difference multiplied by a first weight, the square of the second difference multiplied by a second weight, and the square of the candidate desulfurization dose multiplied by a third weight, as shown in the following formula (1). (1)

[0025] In the above formula (1), J represents the evaluation value, the first weight, the target pH value, the predicted pH value, the second weight, the target sulfur oxide concentration O2, the predicted sulfur oxide concentration, the third weight, and the candidate desulfurization dose. The aforementioned first weight, second weight, and third weight can be adjusted as needed. In these embodiments, the processor 23 selects the candidate desulfurization dose corresponding to the smallest of the evaluation values ​​E_1, ..., E_N as the recommended desulfurization dose. It should be understood that the smaller the square of the first difference corresponding to a candidate desulfurization dose, the closer the pH value of the desulfurization slurry tank T1 will be to the target pH value O1 when the dose of desulfurizing agent added to the desulfurization slurry tank T1 is the candidate desulfurization dose. Furthermore, the smaller the square of the second difference corresponding to a candidate desulfurization dosage, the closer the sulfur oxide concentration at the chimney outlet of the flue gas desulfurization equipment 1 will be to the target sulfur oxide concentration O2 when the dosage of desulfurizing agent added to the desulfurization slurry tank T1 is the candidate desulfurization dosage. Moreover, the lower the candidate desulfurization dosage, the lower the input cost, the less likely the control valve is to be cut and leak, and the less suspended solids are in the wastewater discharged from the flue gas desulfurization equipment, the less likely it is to cause blockage of the wastewater pipeline.

[0026] Establishment of acid-base value prediction model M1 and sulfur oxide concentration prediction model M2

[0027] In some embodiments, the storage device 21 further stores a plurality of historical data sets H_1, ..., H_P of the flue gas desulfurization equipment 1, wherein each of the historical data sets H_1, ..., H_P contains a plurality of historical operating data (not shown) that correspond one-to-one with a plurality of given variables (not shown) of the flue gas desulfurization equipment 1. These given variables of the flue gas desulfurization equipment 1 are variables that are controllable or observable during the operation of the flue gas desulfurization equipment 1.

[0028] In some embodiments, the processor 23 analyzes historical data sets H_1, ..., H_P according to a statistical analysis method (e.g., Pearson correlation coefficient), thereby selecting a complex number of first candidate variables that are more important from the given variables. The processor 23 then determines the characteristic variables based on the first candidate variables and a complex number of second candidate variables (e.g., those selected by process professionals based on experience from the given variables that are more important). In some embodiments, the processor 23 may take the union of the first candidate variables and the second candidate variables as the characteristic variables.

[0029] For ease of understanding, a specific example is provided below, but this specific example is not intended to limit the scope of the invention. In this specific example, the processor 23 selects a first candidate variable from the given variables according to a statistical analysis method, which includes the main steam flow rate, absorber circulation rate, sulfur oxide content at the chimney inlet, flue gas flow rate at the chimney outlet, opening time of control valve P2, and absorber makeup water volume. The second candidate variable selected by a process professional from the given variables includes the main steam flow rate, absorber circulation rate, sulfur oxide content at the chimney inlet, flue gas flow rate at the chimney outlet, opening time of control valve P2, blower air volume, magnesium oxide slurry discharge rate, and magnesium hydroxide slurry temperature. In this specific example, the characteristic variables determined by the processor 23 include the main steam flow rate, absorber circulation rate, sulfur oxide content at the chimney inlet, flue gas flow rate at the chimney outlet, opening time of control valve P2, blower air volume, magnesium hydroxide slurry discharge rate, magnesium hydroxide slurry temperature, and absorber makeup water volume.

[0030] In some embodiments, the processor 23 may also use the first candidate variables as the feature variables, or use the second candidate variables as the feature variables.

[0031] After confirming these characteristic variables, a pH prediction model M1 can be established. Specifically, the historical data set H_1, ..., H_P can be divided into a first subset (not shown) and a second subset (not shown), and the first subset and the second subset have no intersection. The processor 23 trains a first machine learning model based on the first subset of the historical data set H_1, ..., H_P to obtain the pH prediction model M1, and verifies the pH prediction model M1 based on the second subset of the historical data set H_1, ..., H_P. In some embodiments, considering that the operation of the flue gas desulfurization equipment 1 is nonlinear numerical and has time delay characteristics, the first machine learning model can be a nonlinear auto-regressive neural network (NARX).

[0032] It should be noted that the pH prediction model M1 is designed so that its input corresponds to the aforementioned feature variables, and its output corresponds to the predicted pH of the desulfurization slurry tank T1. Those skilled in the art will understand how the processor 23 trains the first machine learning model based on the first subset of historical data set H_1, ..., H_P, which will not be elaborated here. The processor 23 further verifies whether this preliminary pH prediction model M1 meets the requirements based on the second subset of historical data set H_1, ..., H_P. During the verification process, the processor 23 determines whether the root mean square error (RMSE) of the predicted pH value output by the pH prediction model M1 is less than a first threshold value. The aforementioned training and verification can be repeated multiple times until the root mean square error of the predicted pH value output by the pH prediction model M1 is less than the first threshold value.

[0033] After confirming these characteristic variables, a sulfur oxide concentration prediction model M2 can be constructed. Specifically, the historical data set H_1, ..., H_P can be divided into a third subset (not shown) and a fourth subset (not shown), and the third subset and the fourth subset have no intersection. The processor 23 trains a second machine learning model based on the third subset of the historical data set H_1, ..., H_P to obtain the sulfur oxide concentration prediction model M2, and verifies the sulfur oxide concentration prediction model M2 based on the fourth subset of the historical data set H_1, ..., H_P. In some embodiments, considering that the operation of the flue gas desulfurization equipment 1 is nonlinear numerical and has time delay characteristics, the second machine learning model can be a nonlinear autoregressive neural network.

[0034] It should be noted that the sulfur oxide concentration prediction model M2 is designed with its input corresponding to the characteristic variables and the predicted pH value of the desulfurization slurry tank T1, and its output corresponding to the predicted sulfur oxide concentration at the outlet of the flue gas desulfurization equipment 1. Those skilled in the art will understand how the processor 23 trains the second machine learning model based on the third subset of historical data set H_1, ..., H_P, which will not be elaborated here. The processor 23 further verifies whether this preliminary sulfur oxide concentration prediction model M2 meets the requirements based on the fourth subset of historical data set H_1, ..., H_P. During the verification process, the processor 23 determines whether the root mean square error of the predicted sulfur oxide concentration output by the sulfur oxide concentration prediction model M2 is less than a second threshold value. The aforementioned training and verification can be repeated multiple times until the root mean square error of the predicted sulfur oxide concentration output by the predicted sulfur oxide concentration is less than the second threshold value.

[0035] Recommended methods for desulfurization dosage

[0036] The present invention also provides a desulfurization dosage recommendation method (i.e., a method for providing a recommended desulfurization dosage to a flue gas desulfurization device), which is executed by an electronic computing device (e.g., a desulfurization dosage recommendation device 2). The flue gas desulfurization device (e.g., flue gas desulfurization device 1) includes a desulfurization slurry tank and a desulfurization reaction zone. The electronic computing device stores a pH prediction model of the desulfurization slurry tank and a sulfur oxide concentration prediction model of the desulfurization reaction zone. The main flowchart of the desulfurization dosage recommendation method is depicted in Figure 4.

[0037] The desulfurization dosage recommendation method selects a recommended desulfurization dosage from a plurality of candidate desulfurization dosages. For each of the plurality of candidate desulfurization dosages, the desulfurization dosage recommendation method performs steps S401, S403, S405 and S407.

[0038] In step S401, the electronic computing device inputs the candidate desulfurization dose and an operating condition of the flue gas desulfurization equipment into the pH prediction model to obtain a predicted pH value. In step S403, the electronic computing device inputs the candidate desulfurization dose, the operating condition combination, and the predicted pH value into the sulfur oxide concentration prediction model to obtain a predicted sulfur oxide concentration. In step S405, the electronic computing device calculates an evaluation value based on a first difference between the predicted pH value and a target pH value, a second difference between the predicted sulfur oxide concentration and a target sulfur oxide concentration, and the candidate desulfurization dose. In step S407, the electronic computing device determines whether there are any unevaluated candidate desulfurization doses. If the determination result of step S407 is yes, then the process returns to steps S401, S403, and S405 to analyze the unevaluated candidate desulfurization doses. If the determination result of step S407 is negative (i.e., all candidate desulfurization doses have been evaluated), then step S409 is executed. In step S409, the electronic computing device determines the recommended desulfurization dose from the candidate desulfurization doses based on the evaluation values.

[0039] In some embodiments, the evaluation value calculated in step S407 is the sum of the square of the first difference multiplied by a first weight, the square of the second difference multiplied by a second weight, and the square of the candidate desulfurization dose multiplied by a third weight. In these embodiments, step S409 selects the candidate desulfurization dose corresponding to the smallest of these evaluation values ​​as the recommended desulfurization dose.

[0040] In some embodiments, the electronic computing device further stores multiple sets of historical data for the flue gas desulfurization equipment, each set containing multiple sets of historical operating data corresponding one-to-one with multiple given variables. These given variables are controllable or observable variables during the operation of the flue gas desulfurization equipment. In these embodiments, the desulfurization dosage recommendation method further includes a step in which the electronic computing device analyzes the historical data sets according to a statistical analysis method to select multiple first candidate variables from the given variables. The desulfurization dosage recommendation method further includes another step in which the electronic computing device determines the characteristic variables based on the first candidate variables and multiple second candidate variables (e.g., selected by process professionals based on experience from the given variables). In these embodiments, the candidate desulfurization dosage, the combination of operating conditions, and the predicted pH value correspond to multiple characteristic variables.

[0041] In some embodiments, the electronic computing device further stores multiple sets of historical data from the flue gas desulfurization equipment. In these embodiments, the desulfurization dosage recommendation method further includes a step in which the electronic computing device trains a first machine learning model based on a first subset of the historical data sets to obtain the pH prediction model. The desulfurization dosage recommendation method further includes a step in which the electronic computing device verifies the pH prediction model based on a second subset of the historical data sets.

[0042] In some embodiments, the electronic computing device further stores multiple sets of historical data from the flue gas desulfurization equipment. In these embodiments, the desulfurization dosage recommendation method further includes a step in which the electronic computing device trains a second machine learning model based on a third subset of the historical data sets to obtain the sulfur oxide concentration prediction model. The desulfurization dosage recommendation method further includes a step in which the electronic computing device verifies the sulfur oxide concentration prediction model based on a fourth subset of the historical data sets.

[0043] In addition to the above steps, the desulfurization dosage recommendation method provided by the present invention can also perform other steps to have the functions of the desulfurization dosage recommendation device 2 in the aforementioned embodiments, and achieve the same technical effect. Those skilled in the art to which this invention pertains can directly understand how the desulfurization dosage recommendation method provided by the present invention performs these steps based on the aforementioned embodiments, has the same functions, and achieves the same technical effect, so it will not be described in detail.

[0044] The desulfurization dosage recommendation method described in the above embodiments can be implemented by a computer program product containing a plurality of program instructions. The computer program product can be a file that can be transmitted over a network, or it can be stored in a non-transitory computer-readable storage medium. After the program instructions contained in the computer program product are loaded into an electronic computing device (e.g., desulfurization dosage recommendation device 2), the computer program executes the desulfurization dosage recommendation method as described in the above embodiments. The non-transitory computer-readable storage medium can be an electronic product, such as: a read-only memory (ROM), a flash memory, a floppy disk, a hard disk, a compact disk (CD), a digital versatile disc (DVD), a USB flash drive, a database accessible via a network, or any other storage medium known to those skilled in the art and having the same function.

[0045] It should be noted that certain terms in the specification and claims of this invention (including at least: difference, weight, subset, machine learning model, candidate variable) are preceded by "first", "second", "third" or "fourth", which are used to distinguish these terms from each other. Unless otherwise specified, or the order of these terms is not apparent from the context, the order of these terms is not restricted by the prefixes "first", "second", "third" or "fourth".

[0046] In summary, the technology (including at least an apparatus, method, and computer program product) proposed in this invention for providing a suggested desulfurization dosage to a flue gas desulfurization device, in the process of calculating the evaluation value of each candidate desulfurization dosage, first uses an acid-base value prediction model to predict the acid-base value of the desulfurization slurry tank, and then uses a sulfur oxide concentration prediction model to predict the concentration of sulfur oxides at the outlet of the flue gas desulfurization device based on the prediction result of the acid-base value prediction model. By employing the acid-base value prediction model and the acid-base value prediction model executed sequentially, this invention has taken into account the situation where the acid-base value of the desulfurization slurry tank reacts faster after the addition of magnesium hydroxide slurry, while the reaction of the sulfur oxide concentration at the outlet of the flue gas desulfurization device is slower. Furthermore, the present invention takes into account various achievable benefits when calculating the evaluation value of each candidate desulfurization dose, including: the degree of proximity to the target pH value of the desulfurization slurry pool (reflected by the first difference), the degree of proximity to the target sulfur oxide concentration at the outlet of the flue gas desulfurization equipment (reflected by the second difference), and the dosage of the added desulfurizing agent (reflected by the candidate desulfurization dose). Therefore, the present invention can solve all the problems of the prior art.

[0047] The above embodiments are used to illustrate some aspects of the present invention and explain the technical features of the present invention, and are not intended to limit the scope and range of protection of the present invention. Any changes or equivalent arrangements that can be easily made by those skilled in the art to which this invention pertains are within the scope of the present invention, and the scope of protection of the present invention is determined by the claims. [Simplified Explanation of the Diagram]

[0048] Figure 1 depicts a schematic diagram of a typical flue gas desulfurization device 1.

[0049] Figure 2 depicts a schematic diagram of the structure of the desulfurization dosage recommendation device 1 in some embodiments of the present invention.

[0050] Figure 3 depicts the main data flow for providing recommended desulfurization dosage for flue gas desulfurization equipment 1.

[0051] Figure 4 depicts the main flowchart of the desulfurization dosage recommendation method.

Claims

1. An apparatus for providing a recommended desulfurization dose to a flue gas desulfurization (FGD) device, the FGD device comprising a desulfurization slurry tank and a desulfurization reaction zone, the apparatus comprising: a storage device storing an pH prediction model of the desulfurization slurry tank and a sulfur oxide concentration prediction model of the desulfurization reaction zone; and a processor electrically connected to the storage device, and performing the following operations for each of a plurality of candidate desulfurization doses: (a) combining the candidate desulfurization dose with an operating condition of the FGD device and inputting it into the pH prediction model to obtain a predicted pH; (b) inputting the candidate desulfurization dose, the operating condition combination, and the predicted pH into the sulfur oxide concentration prediction model to obtain a predicted sulfur oxide concentration; and (c) calculating an evaluation value based on a first difference between the predicted pH and a target pH, a second difference between the predicted sulfur oxide concentration and a target sulfur oxide concentration, and the candidate desulfurization dose, wherein... The processor further determines the recommended desulfurization dose from among the candidate desulfurization doses based on these evaluation values.

2. The apparatus as claimed in claim 1, wherein the evaluation value calculated by operation (c) is the sum of the square of the first difference multiplied by a first weight, the square of the second difference multiplied by a second weight, and the square of the candidate desulfurization dose multiplied by a third weight, wherein, The processor selects the candidate desulfurization dose corresponding to the smallest of these evaluation values ​​as the recommended desulfurization dose.

3. The apparatus as claimed in claim 1 or 2, wherein the storage further stores a plurality of historical data sets of the flue gas desulfurization equipment, the processor further trains a first machine learning model based on a first subset of the historical data sets to obtain the pH prediction model, and verifies the pH prediction model based on a second subset of the historical data sets.

4. The apparatus as claimed in claim 1 or 2, wherein the storage further stores a plurality of historical data sets of the flue gas desulfurization equipment, the processor further trains a second machine learning model based on a third subset of the historical data sets to obtain the sulfur oxide concentration prediction model, and verifies the sulfur oxide concentration prediction model based on a fourth subset of the historical data sets.

5. The apparatus as claimed in claim 1 or 2, wherein the candidate desulfurization dosage, the combination of operating conditions, and the predicted pH value correspond to a plurality of characteristic variables, wherein, The storage device further stores multiple historical data sets of the flue gas desulfurization equipment. Each historical data set contains multiple historical operating data that correspond one-to-one with multiple given variables. The processor also analyzes these historical data sets according to a statistical analysis method to select multiple first candidate variables from the given variables. The processor also determines the characteristic variables based on the first candidate variables and multiple second candidate variables.

6. A method for providing a recommended desulfurization dose to a flue gas desulfurization (FGD) device, the method being performed by an electronic computing device, the FGD device comprising a desulfurization slurry tank and a desulfurization reaction zone, the electronic computing device storing an acid-base prediction model of the desulfurization slurry tank and a sulfur oxide concentration prediction model of the desulfurization reaction zone, the method comprising the following steps: (a) performing the following steps for each of a plurality of candidate desulfurization doses: (a1) inputting a combination of the candidate desulfurization dose and an operating condition of the FGD device into the acid-base prediction model to obtain a predicted acid-base value; (a2) inputting the candidate desulfurization dose, the combination of operating conditions, and the predicted acid-base value into the sulfur oxide concentration prediction model to obtain a predicted sulfur oxide concentration; and (a3) ​​calculating an evaluation value based on a first difference between the predicted acid-base value and a target acid-base value, a second difference between the predicted sulfur oxide concentration and a target sulfur oxide concentration, and the candidate desulfurization dose; and (b) determining the recommended desulfurization dose from the candidate desulfurization doses based on the evaluation value.

7. The method as described in claim 6, wherein the evaluation value calculated in step (a3) ​​is the sum of the square of the first difference multiplied by a first weight, the square of the second difference multiplied by a second weight, and the square of the candidate desulfurization dose multiplied by a third weight, wherein, Step (b) involves selecting the candidate desulfurization dose corresponding to the smallest of these evaluation values ​​as the recommended desulfurization dose.

8. The method of claim 6 or 7, wherein the electronic computing device further stores a plurality of historical data sets of the flue gas desulfurization equipment, the method further comprising the steps of: training a first machine learning model based on a first subset of the historical data sets to obtain the pH prediction model; and validating the pH prediction model based on a second subset of the historical data sets.

9. The method of claim 6 or 7, wherein the electronic computing device further stores a plurality of historical data sets of the flue gas desulfurization equipment, the method further comprising the steps of: training a second machine learning model based on a third subset of the historical data sets to obtain the sulfur oxide concentration prediction model; and validating the sulfur oxide concentration prediction model based on a fourth subset of the historical data sets.

10. The method of claim 6 or 7, wherein the candidate desulfurization dosage, the combination of operating conditions, and the predicted pH value correspond to a plurality of characteristic variables, the electronic computing device further stores a plurality of historical data sets of the flue gas desulfurization equipment, each of the historical data sets containing a plurality of historical operating data corresponding one-to-one with the plurality of given variables, the method further comprising the following steps: analyzing the historical data sets according to a statistical analysis method, thereby selecting a plurality of first candidate variables from the given variables; and determining the characteristic variables based on the first candidate variables and the plurality of second candidate variables.

11. A computer program product, after being loaded via an electronic computing device, wherein the electronic computing device executes a plurality of program instructions contained in the computer program product to implement the method as described in any one of claims 6 to 10.

Images