Power station boiler NOX emission and efficiency optimization method and system

Abstract

The invention discloses a power station boiler NOx emission and efficiency optimization method and system, and the method comprises the steps: building a mixed model with boiler operation parameters as input and NOx emission and boiler efficiency as output on the basis of working condition experiment data in combination with a least square support vector machine; on the basis of the mixed model, the NOx emission amount and the boiler efficiency serve as targets, boiler operation parameters serve as optimization variables, a multi-target particle swarm algorithm is adopted for optimization, and a boiler NOx emission and efficiency optimization result is obtained. The method has the advantages of high optimization precision and the like.

Description

Technical Field

[0001] The present invention mainly relates to the technical field of boilers, and in particular to a method and system for optimizing NOx emissions and efficiency of power station boilers. Background Art

[0002] For power plants, ensuring safe and economical operation of units, reducing power generation costs, and simultaneously lowering pollutant emissions are essential. Establishing a high-efficiency, low-NOx operation model for boilers based on the current state of power plant boilers, identifying optimal combustion parameters for boilers based on fuel characteristics under varying loads, and achieving high-efficiency, low-NOx combustion in boilers not only has important theoretical significance but also broad application prospects and social benefits.

[0003] To achieve stable and efficient boiler combustion, researchers in the thermal energy field, both domestically and internationally, have been diligently developing effective optimization techniques for boiler combustion. The foundation for guiding safe, stable, and optimized operation of units lies in the ability to establish accurate models. Due to the complexity of modeling boiler combustion characteristics, the development of efficient, low-NOx combustion optimization control systems for boilers is still in its infancy, requiring significant research effort and promising applications. Summary of the Invention

[0004] In view of the technical problems existing in the prior art, the present invention provides a method and system for optimizing NOx emissions and efficiency of power plant boilers with optimized accuracy.

[0005] In order to solve the above technical problems, the technical solution proposed by the present invention is:

[0006] A method for optimizing NOx emissions and efficiency of a power plant boiler, comprising the following steps:

[0007] Based on the experimental data of working conditions and combined with the least squares support vector machine, a hybrid model is established with boiler operating parameters as input and NOx emissions and boiler efficiency as output;

[0008] Based on the hybrid model, with NOx emissions and boiler efficiency as targets and boiler operating parameters as optimization variables, a multi-objective particle swarm optimization algorithm is used to obtain the optimization results of boiler NOx emissions and efficiency.

[0009] Preferably, the boiler operating parameters include one or more of boiler load, carbon received basis content, hydrogen received basis content, nitrogen received basis content, oxygen received basis content, volatile matter received basis content, low calorific value, exhaust gas temperature and furnace outlet oxygen content.

[0010] Preferably, the boiler operating parameters include exhaust gas temperature and furnace outlet oxygen content.

[0011] Preferably, the least squares support vector machine uses a radial basis kernel function as the kernel function, specifically:

[0012] K(x,x i )=exp[||xx i || 2 / 2σ 2 )]

[0013] where x,x i is the input sample, ||xx i || represents the Euclidean distance between samples, σ is the radial basis kernel width parameter, where an increase in σ will make the fitting curve smoother, and a decrease in σ will increase the complexity of the regression function.

[0014] Preferably, the parameter γ that balances the minimization of fitting error and the smoothness of the fitting curve is set to 10, and the width parameter σ of the radial basis kernel function is set to 1. After calculation, the least squares support vector machine model of the relationship between boiler operating parameters and NOx emissions and efficiency is established as follows:

[0015]

[0016] where α = [1.372, 0.881, -0.506, -1.857, 0.444, -0.349, 0.129, 0.038, -1.197, -0.115,

[0017] -0.098, 1.258; -1.110, 1.160, 0.216, -1.428, 0.848, 1.201, -1.405, -0.132, 0.283, -0.461, 0.016, 0.811] T ; b = [-0.012, 0.0286].

[0018] Preferably, the specific process of the multi-objective particle swarm optimization algorithm is:

[0019] Step 1: Initialize the population, calculate the fitness value, and determine the individual optimal solution P based on the fitness value best and the group optimal solution G best ;

[0020] Step 2: Update particle velocity and position according to extreme values;

[0021] Step 3: Calculate the particle fitness value and update the individual optimal solution P according to the fitness value best and the group optimal solution G best ;

[0022] Step 4: Determine whether the loop is finished. If so, return to step 2.

[0023] Preferably, the optimization result of boiler NOx emissions and efficiency is the Parcto optimal solution set, which is a compromise curve between the two objectives of maximum boiler efficiency and minimum NOx emissions.

[0024] The present invention discloses a computer program product, comprising a computer program, which executes the steps of the above method when executed by a processor.

[0025] The present invention also discloses a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the method described above are executed.

[0026] The present invention further discloses a power plant boiler NOx emission and efficiency optimization system, comprising a memory and a processor connected to each other, wherein the memory stores a computer program, and when the computer program is run by the processor, the steps of the above method are executed.

[0027] Compared with the prior art, the advantages of the present invention are:

[0028] The present invention, by understanding the NOx generation mechanism and the calculation formula of the boiler's counter-balanced thermal efficiency, concludes that the furnace outlet excess air coefficient, furnace temperature, exhaust gas temperature, etc. are important influencing factors of boiler NOx emissions and thermal efficiency. The least squares support vector machine (LS-SVM) is used to regress the sample data, and a mixed model of boiler NOx emissions and thermal efficiency is established for the test boiler. The model verification shows that when using the least squares support vector machine, only the coefficient γ and the kernel width σ need to be given to quickly complete its sample training, and the established model has high accuracy and can be used for real-time prediction of NOx emissions and efficiency. It is worth noting that the present invention uses PSO to obtain the optimal value of LS-SVM parameters in a short time (currently there is a difficulty in applying LS-SVM, that is, there is no theoretical guidance for its parameter selection). BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 This is a structural diagram of the hybrid model of the present invention in an embodiment.

[0030] Figure 2 This is the NOx emission regression analysis diagram of the present invention.

[0031] Figure 3 This is a regression analysis diagram of boiler efficiency of the present invention.

[0032] Figure 4 This is the flow chart of the PSO algorithm in the present invention.

[0033] Figure 5 This is the optimization result diagram corresponding to working condition 1 of the present invention.

[0034] Figure 6 This is the optimization result diagram corresponding to working condition 12 of the present invention.

[0035] Figure 7 The flowchart of the optimization method of the present invention in an embodiment. DETAILED DESCRIPTION

[0036] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0037] like Figure 7 As shown, the power plant boiler NOx emission and efficiency optimization method provided by the embodiment of the present invention specifically includes the following steps:

[0038] S1. Based on the experimental data of operating conditions, a hybrid model is established for a specific boiler using the least squares support vector machine (LS-SVM) method. The model takes boiler operating parameters as input and NOx emissions and boiler thermal efficiency as output.

[0039] S2. Based on the hybrid model, with NOx emissions and boiler efficiency as targets and boiler operating parameters (such as exhaust gas temperature and furnace outlet oxygen content) as optimization variables, a multi-objective particle swarm algorithm is used for optimization to obtain the optimization results of boiler NOx emissions and efficiency.

[0040] Through the above optimization, we can achieve both NOx emissions and improved boiler efficiency, so that the comprehensive cost of boiler sewage discharge and fuel consumption is minimized.

[0041] Specifically, the generation of NOx during fuel combustion in power plant boilers is affected by a variety of factors, including boiler design characteristics and boiler operating factors. Specifically:

[0042] (1) Design factors of power plant boilers: These include boiler combustion mode (forward, tangential, front wall), maximum continuous evaporation or boiler capacity, burner type and structure, and heat load of the burner zone;

[0043] (2) Operating parameters of power plant boilers: including oxygen content, load, primary air rate and secondary air distribution method;

[0044] However, since the factors of boiler design are fixed, we will only discuss the impact of power plant boiler operating parameters and fuel characteristics on NOx emissions and boiler efficiency. Specifically:

[0045] 1) Influence of fuel properties

[0046] Since a significant portion of a boiler's NOx emissions comes from nitrogen in the fuel, higher fuel nitrogen content leads to higher NOx emissions. Furthermore, the different forms of nitrogen in the fuel affect NOx generation. Increasing the volatile content of the fuel increases thermal NOx due to earlier ignition points, higher average and peak temperatures, and a corresponding increase in fuel-specific NOx, leading to an increase in total NOx. The ratios of other elements in the fuel's volatile components also affect NOx generation. A higher oxygen-to-nitrogen ratio in the fuel increases NOx emissions. Even at the same oxygen-to-nitrogen ratio, conversion efficiency is affected by factors such as the excess air ratio. A higher excess air ratio results in higher conversion efficiency, leading to increased NOx emissions. Furthermore, the sulfur-to-nitrogen ratio in the fuel also affects emissions, as sulfur and nitrogen compete for oxidation.

[0047] 2) Oxygen content / total air volume / excess air coefficient to NO x The impact of generation

[0048] The excess air coefficient affects both fuel NOx and thermal NOx. When the excess air coefficient is between 0.8 and α and less than 1.1, as α increases, the fuel burns more completely, leading to higher furnace temperatures and a corresponding increase in both fuel NOx and thermal NOx. When α exceeds 1.1, the increase in total air volume with α leads to a decrease in furnace temperature, continuing to increase fuel NOx but decreasing NOx. Generally speaking, NOx emissions increase with increasing α and then level off. In unstaged combustion, reducing the excess air coefficient can reduce NOx emissions by approximately 15% to 20%, but this also increases mechanical losses and incomplete combustion losses, and may cause high-temperature corrosion of the water-cooled walls in the burner area. Therefore, comprehensive consideration should be given to all factors when selecting the excess air coefficient.

[0049] 3) Burnout air to NO x The impact of the amount of production

[0050] The impact of overburn air SOFA / OFA on NOx emissions is significant. When the overburn air volume increases, the air volume of the lower burner will also decrease, which will cause the early combustion of the fuel to be in an oxygen-poor combustion state, and the staged combustion will be more obvious, thereby significantly reducing NOx emissions. Experiments have shown that when the total secondary air volume remains unchanged, when the overburn air rate is increased, the oxygen concentration in the main combustion zone will also decrease. At this time, NOx formation will be significantly suppressed and the denitrification rate will be significantly increased. However, as the overburn air rate increases, the secondary air volume in the main combustion zone will also decrease. Therefore, the first result is the weakening of the wind swirl intensity and the deterioration of the rigidity, which will prevent the formation of a good wind-enclosed fire combustion condition; it may also cause the oxygen concentration in the burner zone to decrease, causing high-temperature corrosion; at the same time, it will cause the fuel to stay too short under high-temperature oxygen-rich conditions, resulting in an increase in the carbon content of the fly ash.

[0051] 4) Load on NO x The impact of generation

[0052] As boiler load increases, the excess air coefficient decreases slightly, leading to a trend toward lower fuel NOx. However, as fuel consumption increases, the furnace temperature rises, significantly increasing thermal NOx emissions. Therefore, total NOx emissions continue to increase with load. Within its normal operating range, the boiler's thermal efficiency also increases with load, but the change is minimal.

[0053] The factors affecting boiler thermal efficiency are:

[0054] When studying the factors that affect boiler efficiency, we must first discuss the thermal balance of the heat medium boiler unit, which will facilitate the analysis of the thermal efficiency of the boiler. Thermal balance is the thermal balance of the heat medium boiler unit when the boiler unit reaches a stable thermal state, with 1kg of liquid or solid fuel or 1m 3 Gas fuel is used as the basis for calculation.

[0055] The corresponding heat balance equation for each kg of fuel is as follows:

[0056] Q r =Q1+Q2+Q3+Q4+Q5+Q6

[0057] Expressed as a percentage of heat input:

[0058] 100% = q1+q2+q3+q4+q5+q6

[0059] In the above formula, Q1 is the heat input to the boiler (kJ / kg), Q2 is the heat effectively utilized by the boiler, and Q1 is the exhaust heat loss.

[0060] Q3 is the heat loss due to incomplete combustion of gas, Q4 is the heat loss due to incomplete combustion of solid, Q5 is the heat loss due to heat dissipation, and Q6 is the physical heat loss due to ash; q1, q2, q3, q4, q5, and q6 are the percentages corresponding to Q1, Q2, Q3, Q4, Q5, and Q6 respectively.

[0061] The percentage of heat effectively utilized by the boiler to the total heat input is the boiler efficiency η, that is:

[0062]

[0063] In the test, the heat input Q is measured r The positive balance method is used to calculate the boiler efficiency by effectively utilizing the heat Q1. For boilers, air preheaters are usually used to preheat the air, so the input heat Q r Right now:

[0064]

[0065] In the formula is the low calorific value (kJ / kg), i r The physical heat brought by the fuel. When the fuel moisture is relatively low, the above formula can be simplified to:

[0066]

[0067] For power plant boilers, it is very difficult to measure the amount of fuel when determining the input heat, and it often causes large errors when measuring the output and input heat. At the same time, the positive balance method only calculates the efficiency of the boiler unit, but does not measure the various heat losses. Therefore, it is difficult to analyze the causes of various heat losses and find ways to reduce heat losses. Therefore, in practice, the reverse balance method is used to calculate the thermal efficiency of the boiler unit. The calculation formula of the reverse balance method is:

[0068] η=100%-(q2+q3+q4+q5+q6)

[0069] The q2, q3, q4, q5, and q6 in the formula jointly determine the thermal efficiency of the boiler.

[0070] Furthermore, the combustion process in power plant boilers is a complex chemical and physical process. Numerous factors influence this process, and it exhibits nonlinear and coupled characteristics. This complexity is difficult to describe using a mechanistic model. However, because support vector machines are black-box models, the nonlinear functional relationship between their input and output can be implemented using support vector machines, making them suitable for modeling boiler combustion response characteristics.

[0071] Specifically, LS-SVM is used to establish a boiler combustion regression prediction model (hybrid model). The actual measured data of boiler operation is used as the training set, and various parameters affecting boiler combustion are used as input, and the boiler NOx emissions or thermal efficiency are used as output. Then, appropriate LS-SVM parameters are selected, and the Lagrange multiplier a and bias b are calculated through training to find the causal relationship between boiler operating parameters and NOx emissions or thermal efficiency.

[0072] The boiler combustion prediction model of LS-SVM is as follows: Figure 1 shown. Figure 1 In the system, the adjustable input parameters are concentrated in the boiler operation. The operator can adjust the parameters by changing the set values in the DCS, including the oxygen content at the boiler furnace outlet, the air distribution method, etc. For some working conditions, the boiler load is a non-adjustable parameter. i (i=1,2,…m) is the m training samples that constitute the input parameters.

[0073] Due to practical limitations, all training data was collected from power plant fuel composition analysis and DCS data, which has high collection accuracy. The collected data is shown in Tables 1 and 2.

[0074] Table 1 Operating parameters of a power plant boiler and NO x Emissions and thermal efficiency

[0075]

[0076] Table 2 Operating parameters of a power plant boiler and NO x Emissions and thermal efficiency

[0077]

[0078]

[0079] The specific process of establishing the hybrid model in step S1 is:

[0080] (1) Model structure

[0081] Many factors influence boiler efficiency and NOx emissions. NOx is generated primarily through three pathways: rapid, fuel, and thermal. Rapid NOx generation is minimal, with the latter two factors having the greatest impact on NOx generation. Fuel-type NOx generation primarily depends on the air distribution method and nitrogen content in the fuel. Thermal NOx generation primarily depends on temperature and nitrogen content. Therefore, boiler load, carbon base content, hydrogen base content, nitrogen base content, oxygen base content, volatile base content, low-level heating value, flue gas temperature, and furnace outlet oxygen content are selected as inputs affecting NOx emissions and thermal efficiency, with NOx emissions and thermal efficiency serving as outputs.

[0082] (2) Selection of LS-SVM parameters

[0083] LS-SVM converts nonlinear transformations into inner product operations in a high-dimensional feature space into kernel function calculations in the original space. This significantly reduces the amount of computation required to avoid direct operations in the high-dimensional feature space and avoids the "curse of dimensionality" in artificial neural network algorithms. Furthermore, after using the kernel function, it is even necessary to know the specific form of the nonlinear function. LS-SVM uses the radial basis kernel function in numerical simulations:

[0084] K(x,x i )=exp[||xx i || 2 / 2σ 2 )]

[0085] This is mainly because when there is no prior knowledge about the problem, the model trained by this kernel function will have better overall performance than other kernel function models.

[0086] From the derivation of the least squares support vector machine algorithm, we can see that the main parameters of the least squares support vector machine using the radial basis kernel are the radial basis kernel parameter σ and the hyperparameter γ. These two parameters can largely determine the prediction and learning capabilities of the least squares support vector machine. γ is a parameter that minimizes the fitting error and balances the smoothness of the fitting curve. When the γ value is larger, the penalty for training error will also increase, and the point fitting values of the training data samples will be more consistent with the actual values, but it is also prone to overfitting. Reducing the γ value will reduce the complexity of the model; increasing σ will make the fitting curve smoother, while decreasing the σ value will increase the complexity of the regression function and also easily cause overfitting.

[0087] SVM parameter selection lacks theoretical guidance. Currently, parameter selection is typically based on a trial-and-error approach, whereby a satisfactory solution is manually selected. This method relies on subjective experience and is time-consuming.

[0088] (3) Power plant boiler NO x Hybrid model of emissions and thermal efficiency

[0089] The parameter γ, which balances the minimization of fitting error and the smoothness of the fitting curve, is set to 10, and the width parameter σ of the radial basis kernel function is set to 1. After calculation, the least squares support vector machine model of the relationship between boiler operating parameters and NOx emissions and efficiency is established as follows:

[0090]

[0091] where α = [1.372, 0.881, -0.506, -1.857, 0.444, -0.349, 0.129, 0.038, -1.197, -0.115,

[0092] -0.098, 1.258; -1.110, 1.160, 0.216, -1.428, 0.848, 1.201, -1.405, -0.132, 0.283, -0.461, 0.016, 0.811] T ; b = [-0.012, 0.0286].

[0093] In order to verify the correctness of the model, a linear regression analysis of the actual value and the model estimated value was performed on the training samples of NOx emissions and efficiency. Figure 2 、 Figure 3 The linear regression correlation coefficient, R = I, demonstrates that the support vector machine fits the training samples very accurately. Generalization testing of the trained model using test samples revealed relative errors in NOx emissions and efficiency of no more than 0.5%, as shown in Table 3. This model demonstrates excellent generalization and is well suited for identifying boiler NOx emissions and thermal efficiency models. This support vector machine can be used as a boiler prediction model in practical applications.

[0094] Table 3 Comparison of generalization ability between LS-SVM and MARN models

[0095]

[0096] The prediction model inputs include both non-adjustable and adjustable parameters. Non-adjustable parameters primarily include the layout and structure of the boiler burner and the furnace structure. These require hardware optimization through equipment modification and other methods. Boiler load is also a non-adjustable parameter for the operating conditions that require optimization. Adjustable input parameters primarily focus on parameters that the boiler operator can adjust by changing setpoints in the DCS. These include total air volume and the boiler's air distribution method.

[0097] Based on the principle of controllable operating variables, the exhaust gas temperature and furnace outlet oxygen content in the model input are selected as optimization variables in this embodiment. Considering safety, the exhaust gas temperature range is 150℃ to 186℃, and the furnace outlet oxygen content range is 3.3% to 5.1%.

[0098] Based on the established model, in the MATLAB environment, the first and twelfth groups of operating conditions with the most serious NOx emissions in the samples were selected for optimization calculation. The MOPSOCD parameters were selected as: population size 100, iteration number 30. The optimization results are as follows: Figure 5 and Figure 6 As shown. Figure 5 It can be seen that after optimization, the NOx emission concentration has dropped significantly and the boiler efficiency has been improved. This shows that there is a large energy-saving potential and optimization space in operating condition 1. The reduction of NOx emissions and the improvement of boiler thermal efficiency are not contradictory within the scope of the optimization space. Figure 6 In the case of a boiler, thermal efficiency does not improve, but instead decreases as NOx emission concentration decreases. The two trends of decreasing and increasing together indicate a conflict between the two. Therefore, when the potential for optimizing boiler operating conditions is low or nonexistent, it is inadvisable to unilaterally emphasize either reducing NOx emissions or increasing boiler efficiency.

[0099] It's important to note that in this implementation, the solutions to the multi-objective optimization problem are Parceto-optimal solutions, not global optimal solutions. These solutions are essentially a compromise curve between maximizing boiler efficiency and minimizing NOx emissions, which is the fundamental concept of MOPSO. From the Pareto distribution curve, boiler operators can easily select the ideal operating point based on their experience and the importance of the objectives.

[0100] Select any one solution from the Pareto solution set to compare working condition 1 and working condition 12 before and after optimization, and the results are shown in Table 4. For working condition 1, the exhaust gas temperature decreases after optimization, while the flue gas oxygen content increases. When the flue gas oxygen content is 5.08% and the exhaust gas temperature is 155°C, the NOx emission concentration is 177.5mg / m3, and the boiler efficiency is 89.8%. Analysis shows that: the reduction of exhaust gas temperature and the reduction of exhaust gas heat loss will improve the thermal efficiency of the boiler; and the increase in exhaust gas oxygen content promotes combustion, but also increases NOx emissions. For working condition 12, the exhaust gas temperature and remain basically unchanged after optimization, which reduces the NOx emissions from the original 210mg / m 3 Reduced to 190.2 mg / m 3 The optimized furnace outlet oxygen volume is also increased, which is conducive to the complete combustion of fuel and reduces combustion losses.

[0101] Table 4 Comparison before and after optimization

[0102]

[0103] The above multi-objective particle swarm algorithm only needs to give the population size and the number of iterations, and a set of Pareto optimal solutions can be obtained in one run.

[0104] Related explanation: The PSO algorithm searches for the optimal solution in a complex space through competition and collaboration between individuals, while individuals follow certain rules of movement. The PSO algorithm initializes a group of particles in the solvable space. Each particle represents a potential optimal solution to the extreme optimization problem. Each particle is represented by three indicators: position, speed, and fitness value. The fitness value is determined by the fitness function and is a quantity that characterizes the quality of the particle position. The particles move in the solution space and track the optimal solution of the group G. best and the individual optimal solution P best To update the individual position, each time the position is updated, its fitness value is calculated once, and the group extreme value and individual extreme value positions are updated by comparing the new particle fitness value with the fitness value of the group extreme value and the individual extreme value.

[0105] like Figure 4 As shown in Figure 2, the specific steps of the PSO algorithm are:

[0106] Step 1: Initialize the population, calculate the fitness value, and determine the individual optimal solution P based on the fitness value best and the group optimal solution G best ;

[0107] Step 2: Update particle velocity and position according to extreme values;

[0108] Step 3: Calculate the particle fitness value and update the individual optimal solution P according to the fitness value best and the group optimal solution G best ;

[0109] Step 4: Determine whether the loop is finished. If so, return to step 2.

[0110] The present invention, by understanding the NOx generation mechanism and the calculation formula of the boiler's counter-balanced thermal efficiency, concludes that the furnace outlet excess air coefficient, furnace temperature, exhaust gas temperature, etc. are important influencing factors of boiler NOx emissions and thermal efficiency. The least squares support vector machine (LS-SVM) is used to regress the sample data, and a mixed model of boiler NOx emissions and thermal efficiency is established for the test boiler. The model verification shows that when using the least squares support vector machine, only the coefficient γ and the kernel width σ need to be given to quickly complete its sample training, and the established model has high accuracy and can be used for real-time prediction of NOx emissions and efficiency. It is worth noting that the present invention can obtain the optimal value of the LS-SVM parameters in a short time by using the PSO program (currently there is a difficulty in applying LS-SVM, that is, there is no theoretical guidance for its parameter selection).

[0111] The essence of the present invention is to improve the thermal efficiency of the boiler on the basis of reducing NOx emissions, which is a multi-objective optimization problem. Based on the established model, a multi-objective particle swarm optimization algorithm (MOPSO) is tried to optimize the adjustable operating parameters of the boiler. The results show that this algorithm only needs to give two parameters, the number of iterations and the population size, to obtain the Pareto optimal solution set through a single run. The optimization calculations for the two operating conditions show that reducing NOx emissions and improving the thermal efficiency of the boiler are not contradictory in all cases. Only when the optimization potential of the operating conditions is small or there is no optimization space, reducing NOx emissions requires sacrificing boiler efficiency, and one-sided emphasis cannot be placed on reducing NOx emissions or improving boiler efficiency.

[0112] The present invention discloses a computer program product, including a computer program that, when executed by a processor, performs the steps of the method described above. The present invention also discloses a computer-readable storage medium having a computer program stored thereon that, when executed by a processor, performs the steps of the method described above. The present invention further discloses a power plant boiler NOx emission and efficiency optimization system, including an interconnected memory and a processor, the memory storing a computer program that, when executed by the processor, performs the steps of the method described above. The products, media, and systems of the present invention correspond to the method described above and share the advantages described by the method described above.

[0113] The present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by hardware related to computer program instructions. The computer program can be stored in a computer-readable storage medium, and when the computer program is executed by the processor, it can implement the steps of the above-mentioned method embodiment. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. Computer-readable storage media include: any entity or device that can carry computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc. The memory is used to store computer programs and / or modules, and the processor implements various functions by running or executing computer programs and / or modules stored in the memory, and calling data stored in the memory. The memory may include high-speed random access memory and non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage device.

[0114] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions based on the principles of the present invention are within the scope of protection of the present invention. It should be noted that for those skilled in the art, various improvements and modifications that do not depart from the principles of the present invention should be considered within the scope of protection of the present invention.

Claims

1. A method for optimizing NOx emissions and efficiency of power plant boilers, characterized in that: Including steps: Based on the experimental data of working conditions and combined with the least squares support vector machine, a hybrid model is established with boiler operating parameters as input and NOx emissions and boiler efficiency as output; Based on the hybrid model, with NOx emissions and boiler efficiency as targets and boiler operating parameters as optimization variables, a multi-objective particle swarm optimization algorithm is used to obtain the optimization results of boiler NOx emissions and efficiency.

2. The power plant boiler NOx emission and efficiency optimization method according to claim 1, characterized in that: The boiler operating parameters include one or more of boiler load, carbon received basis content, hydrogen received basis content, nitrogen received basis content, oxygen received basis content, volatile matter received basis content, low calorific value, exhaust gas temperature and furnace outlet oxygen content.

3. The power plant boiler NOx emission and efficiency optimization method according to claim 2, characterized in that: The boiler operating parameters include exhaust gas temperature and furnace outlet oxygen content.

4. The power plant boiler NOx emission and efficiency optimization method according to claim 1, 2 or 3, characterized in that: The least squares support vector machine uses the radial basis kernel function as the kernel function, specifically: K(x,x i )=exp[||x-x i || 2 / 2σ 2 )] where x,x i is the input sample, ||xx i || represents the Euclidean distance between samples, and σ is the radial basis kernel width parameter. Increasing σ will make the fitting curve smoother, while decreasing σ will increase the complexity of the regression function.

5. The power plant boiler NOx emission and efficiency optimization method according to claim 4, characterized in that: The parameter γ, which balances the minimization of fitting error and the smoothness of the fitting curve, is set to 10, and the width parameter σ of the radial basis kernel function is set to 1. After calculation, the least squares support vector machine model of the relationship between boiler operating parameters and NOx emissions and efficiency is established as follows: where α = [1.372, 0.881, -0.506, -1.857, 0.444, -0.349, 0.129, 0.038, -1.197, -0.115, -0.098, 1.258; -1.110, 1.160, 0.216, -1.428, 0.848, 1.201, -1.405, -0.132, 0.283, -0.461, 0.016, 0.811] T ; b = [-0.012, 0.0286].

6. The power plant boiler NOx emission and efficiency optimization method according to claim 1, 2 or 3, characterized in that: The specific process of the multi-objective particle swarm optimization algorithm is as follows: Step 1: Initialize the population, calculate the fitness value, and determine the individual optimal solution P based on the fitness value best and the group optimal solution G best ; Step 2: Update particle velocity and position according to extreme values; Step 3: Calculate the particle fitness value and update the individual optimal solution P according to the fitness value best and the group optimal solution G best ; Step 4: Determine whether the loop is finished. If so, return to step 2.

7. The power plant boiler NOx emission and efficiency optimization method according to claim 1, 2 or 3, characterized in that: The optimization results of boiler NOx emissions and efficiency are the Parcto optimal solution set, which is a compromise curve between the two objectives of maximum boiler efficiency and minimum NOx emissions.

8. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are performed.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the computer program performs the steps of the method according to any one of claims 1 to 7.

10. A power plant boiler NOx emission and efficiency optimization system, comprising a memory and a processor connected to each other, wherein a computer program is stored in the memory, characterized in that: When the computer program is executed by a processor, the computer program performs the steps of the method according to any one of claims 1 to 7.

Images