Fuel Combination Optimization Model of Thermal Power Plant Based on New Particle Swarm Optimization Algorithm

Influenced by the unbalanced state of particle swarm in the process of fuel combustion in thermal power plants, the fuel cost in the thermal power generation stage is relatively high. Therefore, a new particle swarm optimization model for fuel combination in thermal power plants is proposed. Combined with the combustion properties of different fuels, from the point of view of particle swarm optimization, in the process of carrying out specific particle swarm optimization simulation, the original particle swarm optimization algorithm is improved adaptively. A new particle swarm optimization algorithm is constructed by coupling it with a multiphase turbulence model. The fuel combustion performance of the power plant is analyzed by using its coupling. The optimization model of fuel combination in the thermal power plant is constructed by maximizing the energy release of fuel as the core, and the state of the fluid-particle field is taken as the constraint condition. In the test results, the design of the fuel combination optimization model can fully improve the energy release degree of fuel and reduce fuel consumption under the same power generation demand, which has a positive effect on power generation cost control.


Introduction
For thermal power plants, one of the most important cost inputs is the fuel cost in the power generation stage.However, due to the influence of the objective combustion state, there is a situation in which the energy release of some fuels is low [1][2][3], which greatly affects the effective utilization value of fuel and increases the operating cost of power enterprises to a certain extent.Therefore, more and more attention has been paid to the research based on the fuel combination of thermal power plants.Among them, the real-time optimization model of DME/LPG fuel ratio with HCCI combustion control as the core is one of the more common application models.It not only designs the fuel composition finely (DME/LPG mixed fuel) [4][5][6] but also conducts targeted research on the mixture composition in the combustion environment (adding carbon dioxide into the intake air) and adjusts the engine parameters (changing the compression ratio).Thus, it can realize the optimization of fuel ratio under different working conditions and improve the combustion characteristics of fuel.However, this model is based on the existing fuel resources and lacks the utilization and analysis of new fuels.In addition, the COREX fuel proportioning model based on decision tree algorithm rule extraction is also a widely used model at present [7][8][9], and its specific research object is the COREX-3000-2# boiler.When learning the sample data, the decision tree C4 5 rule extraction mechanism is introduced, and the quantitative relationship between fuel proportioning and output coke ratio [10][11] is analyzed based on knowledge rules, which greatly improves the energy release of fuel [12][13].
On the basis of the above, this paper puts forward the research on the fuel combination optimization model of thermal power plants based on the new particle swarm optimization algorithm.It designs the comparative test environment and analyzes and verifies the application value of the model designed in this paper.

Coupling analysis of fuel combustion performance of power plant based on new particle swarm optimization algorithm
In order to realize the reasonable optimization of fuel combinations in thermal power plants, it is extremely necessary to comprehensively analyze different fuels from the point of view of particle swarm balance.In view of this, from the perspective of a multi-dimensional space system, this paper carried out the simulation study of particle swarm equilibrium so as to realize the coupling analysis of fuel combustion performance in power plants [14][15].Among them, in the specific numerical simulation analysis of a fuel particle swarm, it should have at least three basic characteristics.Firstly, the data parameters of the fuel particle swarm can be coupled with the numerical algorithm describing the two-phase turbulent field, and it has extended properties.Secondly, the numerical value of the fuel particle swarm can be introduced into the grid division algorithm.It also shows the expansion performance.Finally, considering the four-way coupling, the fuel particle swarm value can also support the expansion.This means that on the basis of considering various dynamic events, the influence of the fuel combustion environment on the particle field and fluid field in thermal power plants exists objectively and can be estimated separately.Correspondingly, the interference caused by the above influence on the dynamic evolution process can also be reflected in quantitative form.Then, at this time, it is necessary to couple particle swarm equilibrium simulation with continuous-phase turbulence simulation so as to describe the dynamic evolution characteristics of particle size distribution by particle swarm equilibrium simulation.Combined with the above theoretical basis, this paper fully considers the complexity of the thermal power plant in the process of dynamic evolution.In order to describe the development relationship of fuel particle swarm more accurately and ensure the higher integrity of the fuel combination optimization model in the final design of the thermal power plant, in the process of carrying out the specific particle swarm balance simulation, the adaptive improvement is made.It is coupled with the multiphase turbulence model, and a new particle swarm optimization algorithm is constructed.The specific implementation process is shown in Figure 1.Coupling analysis process of fuel combustion performance in power plant based on new particle swarm optimization algorithm.According to the above-mentioned way, the velocity of the fluid micro-cluster seen on the particle track is determined, and the intensity and state of turbulent interaction between two phases are defined at the same time.The fluid field and particle field are exchanged in time step, and the state of the fluid-particle field in the combustion furnace of the four-way coupling thermal power plant is realized so as to realize the simulation of specific values.

Construction of fuel combination optimization model for thermal power plant
Combined with the coupling analysis results of the fluid-particle field state in the combustion furnace of a thermal power plant in part 1.1, this paper focuses on maximizing the energy release of fuel when designing the fuel combination optimization model.Among them, the specific optimization model can be expressed as: max among them, max z represents the maximum total energy release of fuel combination in thermal power plant, i c represents that instantaneous velocity parameter of the fluid micelle seen by i fuel particle, i x represents the drift coefficient of I fuel particles in phase space, i p represents the energy reduction coefficient considering the "track crossing effect", "continuity effect" and "inertia effect" of i fuel particles, and b indicates the slow term parameter of fuel in thermal power plant.
It should be noted that the fuel combination optimization model shown in Formula (1) needs to be carried out under certain constraints; that is, the state of the fluid-particle field is within the allowable fluctuation range, in combination with the actual demand of fuel particle group balance in thermal power plants.In view of this, this paper constructs the constraint function of the fuel combination optimization model, which can be expressed as: among them, v represents the probability that i fuel particles collide/coalesce with other virtual particles,  indicates the collision/coagulation coefficient of particles in the fuel combustion environment of a thermal pow plant, and min i w represents the minimum fluid phase motion decoupling of i fuel particles.According to the above-mentioned way, the fuel combination optimization model of the thermal power plant is constructed.

Test environment
When analyzing the application performance of the fuel combination optimization model of the thermal power plant designed in this paper based on the new particle swarm optimization algorithm, in order to objectively evaluate its application value intuitively, a comparative test was carried out.Among them, the specific test environment is an actual thermal power plant, in which it is necessary to make a coal blending list with a term of 8 hours.On this basis, the load of the generator set in the test environment was analyzed, specifically 400, 000 kilowatts, and was kept stable for 8 consecutive hours.During the fuel testing process, coal sample analysis instruments were used to sample and analyze each fuel, measuring parameters such as sulfur content and lower calorific value of the coal sample.The calorific value of each fuel was measured using a Harris Karl calorimeter, and the energy consumption of the fuel was determined using a Smith calorimeter.The composition and properties of coal samples were analyzed using the Leco CS elemental analyzer.Real-time monitoring of environmental parameters was conducted, such as concentration and temperature of combustion emissions, through the Testo 340 multifunctional flue gas analyzer and recording of corresponding data.According to environmental requirements, the CEMS system was used to monitor and record the unit time emissions of combustion emissions during fuel testing, ensuring that they do not exceed the specified limits.The available fuel resources were counted, including four kinds, and the related parameter information is shown in Table 1.
Table 1 Based on the available fuel (raw coal) resource parameter information shown in Table 1, the coal consumption is 350 kWh according to the coal consumption-load relationship of thermal power plants standard to carry out the test.Among them, from the perspective of environmental constraints, the sulfur emission per hour should be controlled within 1.08 tons; that is, the emission per unit time should not exceed 300 g.In the specific test process, this paper sets up a comparative test environment in which the control groups participating in the test are DME/LPG fuel ratio real-time optimization model and fuel ratio model based on decision tree algorithm rule extraction.By comparing the test results under different models, the application value of the model designed in this paper is objectively evaluated.

Test results and analysis
On the basis of the above test environment, this paper takes the fuel cost input under different load conditions as the evaluation index, and the specific test results are shown in Figure 2. Based on the analysis of the application effects of the three models under different load conditions shown in Figure 2, it can be seen that when the load of the test environment generator set is 400, 000 kWh, the fuel cost input corresponding to the DME/LPG fuel ratio real-time optimization model is the highest.When the load of the test environment generator set reaches more than 400, 000 kW, the fuel cost input of the decision tree algorithm is the highest.In contrast, the cost of the fuel combination optimization model designed in this paper under different load conditions is obviously lower than that of the control group.Among them, when the load of the generator set in the test environment is 400, 000 kWh, the fuel cost input is 113.1 yuan lower than that of the DME/LPG fuel ratio real-time optimization model and 100.0 yuan lower than that of the decision tree algorithm rule extraction fuel ratio model.When the load of the generator set in the test environment is 500, 000 kWh, the fuel cost input is 130.5 yuan lower than that of the DME/LPG fuel ratio real-time optimization model and 137.3 yuan lower than that of the decision tree algorithm.When the load of the generator set in the test environment is 600, 000 kWh, the fuel cost input is 37.9 yuan lower than that of the DME/LPG fuel ratio real-time optimization model and 71.6 yuan lower than that of the decision tree algorithm rule extraction fuel ratio model.Combined with the above test results, it can be concluded that the fuel combination optimization model of the thermal power plant designed in this paper based on the new particle swarm optimization algorithm can fully improve the energy release degree of fuel and reduce

Figure 1 .
Figure 1.Coupling analysis process of fuel combustion performance in power plant based on new particle swarm optimization algorithm.According to the above-mentioned way, the velocity of the fluid micro-cluster seen on the particle track is determined, and the intensity and state of turbulent interaction between two phases are defined at the same time.The fluid field and particle field are exchanged in time step, and the state of the fluid-particle field in the combustion furnace of the four-way coupling thermal power plant is realized so as to realize the simulation of specific values.

Figure 2 .
Figure 2. Comparison chart of fuel cost input.Based on the analysis of the application effects of the three models under different load conditions shown in Figure2, it can be seen that when the load of the test environment generator set is 400, 000 kWh, the fuel cost input corresponding to the DME/LPG fuel ratio real-time optimization model is the highest.When the load of the test environment generator set reaches more than 400, 000 kW, the fuel cost input of the decision tree algorithm is the highest.In contrast, the cost of the fuel combination optimization model designed in this paper under different load conditions is obviously lower than that of the control group.Among them, when the load of the generator set in the test environment is 400, 000 kWh, the fuel cost input is 113.1 yuan lower than that of the DME/LPG fuel ratio real-time optimization model and 100.0 yuan lower than that of the decision tree algorithm rule extraction fuel ratio model.When the load of the generator set in the test environment is 500, 000 kWh, the fuel cost input is 130.5 yuan lower than that of the DME/LPG fuel ratio real-time optimization model and 137.3 yuan lower than that of the decision tree algorithm.When the load of the generator set in the test environment is 600, 000 kWh, the fuel cost input is 37.9 yuan lower than that of the DME/LPG fuel ratio real-time optimization model and 71.6 yuan lower than that of the decision tree algorithm rule extraction fuel ratio model.Combined with the above test results, it can be concluded that the fuel combination optimization model of the thermal power plant designed in this paper based on the new particle swarm optimization algorithm can fully improve the energy release degree of fuel and reduce

.
Statistical table of available fuel (raw coal) resource parameter information.