Parameter correction method for full-process simulation model of MSWI based on adaptive genetic algorithm

Abstract

The present application relates to a MSWI full process simulation model parameter correction method based on an adaptive genetic algorithm, comprising: obtaining a MSWI full process numerical simulation model, determining the chemical reaction equilibrium and phase equilibrium of the RGibbs module in the MSWI full process numerical simulation model; performing parameter analysis on the chemical reaction equilibrium and phase equilibrium, determining the temperature set value as a correction parameter, and taking the gas minimization at G3 in the MSWI process as an objective function, iteratively solving the correction parameter by an adaptive genetic algorithm to obtain an optimal temperature set value. The present application improves the precision of model simulation of various garbage incineration plant working conditions, reduces the deviation of simulation results and actual data, and provides support for understanding the incineration performance under different parameters and the construction of intelligent garbage incineration plants.

Description

Technical Field

[0001] This invention relates to the field of information technology, and in particular to a method for parameter correction of the MSWI full-process simulation model based on an adaptive genetic algorithm. Background Technology

[0002] Numerical simulation models are a feasible means of studying the characteristics of MSWI processes and have been widely used in MSWI research. For example, existing technical solutions have established waste incineration models that can analyze grate speed and primary / secondary air ratios, analyzing the changes under different grate speeds and air ratios. The results show that under G1 conditions, the error fluctuation ranges of O2, CO2, SO2, SO3, NO, and NO2 are 5.9%–48.5%, 2.5%–11.5%, 10.42%–28.57%, 7%–23%, 29.10%–45.52%, and 24.39%–48.78%, respectively. Existing technical solutions also disclose the establishment of a waste pyrolysis gasification model based on Aspen Plus, and sensitivity analysis of key operating parameters and overall efficiency. Existing technical solutions also disclose a simulation model based on an established 600t / d incinerator, which analyzes in detail the combustion characteristics of the bed and furnace under different moisture contents and inlet temperatures. The results show that when the inlet temperature is 30℃, the volatile matter release rate and fixed carbon combustion rate are negatively correlated with the moisture content. This result provides useful information for the pretreatment of municipal solid waste. Existing technologies also establish a model for predicting CO emissions. This modeling method integrates virtual and real data. Using virtual data as the dataset, a mechanistic mapping model is established using a linear regression decision tree algorithm; a process mapping model is established using real data; then, the two models are fused with expert experience. The results show that this method can effectively predict CO emissions. The parameter settings of the simulation model are a key factor affecting the accuracy and reliability of the results; improper settings can easily lead to large deviations between the simulation results and actual values. However, the simulation strategy used in the above studies did not consider the adjustment of parameter settings.

[0003] With the continuous development of artificial intelligence, machine learning algorithms are being applied more and more widely in industrial processes. For example, to improve the prediction accuracy of battery simulation models, existing technologies have precisely selected key model parameters. To reduce the difference between simulated and actual battery voltage values, battlefield optimization algorithms are used to determine model parameter values. Results show that optimizing the key parameters of the battery model significantly improves its prediction accuracy. Existing technologies propose a model parameter calibration framework, using Janselman derivative techniques to optimize model parameters calculated from phase diagrams. Experimental results show that this method can effectively calculate the gradient of thermodynamic properties with respect to model parameters under equilibrium conditions. Existing technologies also utilize optimization algorithms to study the correction of multiple model parameters to improve the accuracy of nuclear power plant simulation models, employing particle swarm optimization to optimize subsystems in the entire process model. Experimental results show that the error of the optimized simulation model is less than 2%, providing support for adaptive model adjustment and enhancing its practical application value. Existing technologies also propose a dual-window concept drift detection method for industrial process parameter modeling, solving the performance degradation of soft measurement models and model update problems caused by the inherent concept drift characteristics of data. Existing technical solutions also address the problem that octane number harmonization models struggle to simulate different operating conditions. A parameter update method based on recursive least squares is proposed, demonstrating that this method can quickly adapt to simulations under different operating conditions, thus supporting improved production efficiency. Furthermore, existing technical solutions utilize a genetic algorithm (GA) to estimate the reaction kinetics and runoff kinetics parameters of the acetylene hydrogenation reaction system. Experimental results show that the tasks of each reactor are reasonably allocated, further improving the unit's efficiency. GA is a global optimization search method that does not require explicit mathematical equations or derivative expressions of the objective function, offering advantages such as simplicity, efficiency, and resistance to getting trapped in local optima. It has successfully solved problems related to wiring schemes, parameter estimation, cognitive models, combinatorial optimization, and adaptive control.

[0004] Although existing technical solutions have made significant progress, some issues still require further research: 1) The operating conditions of waste incineration plants fluctuate significantly, and the nature of different operating conditions also varies. However, simulation models typically use fixed parameter settings, leading to inaccurate simulations.

[0005] 2) The simulation model of waste incineration power generation involves a variety of physical properties, flow and geometric parameters. Existing technical solutions do not analyze the impact of different characteristic parameters on the results.

[0006] 3) The numerical simulation model for waste incineration power generation and the artificial intelligence algorithm run on different software, and it is difficult for different software to interact directly. Summary of the Invention

[0007] To address the problems existing in the prior art, the purpose of this invention is to provide a parameter correction method for the MSWI full-process simulation model based on an adaptive genetic algorithm. This method improves the accuracy of the model in simulating various operating conditions of waste incineration plants, reduces the deviation between simulation results and actual data, and provides support for understanding the incineration performance under different parameters and the construction of intelligent waste incineration plants.

[0008] To achieve the above objectives, the present invention provides the following solution: A parameter calibration method for the MSWI full-process simulation model based on adaptive genetic algorithm includes: Obtain the MSWI full-process numerical simulation model and determine the chemical reaction equilibrium and phase equilibrium of the RGibbs module in the MSWI full-process numerical simulation model; The chemical reaction equilibrium and the phase equilibrium are analyzed by parameters to determine the temperature setpoint as the correction parameter. The optimal temperature setpoint is obtained by iteratively solving the correction parameter using an adaptive genetic algorithm with the goal of minimizing the gas at G3 in the MSWI process.

[0009] Optionally, the chemical reaction equilibrium includes: ; ; ; in, The free energy of the system after the reaction reaches thermal equilibrium. Reaching the minimum value, As a phase that exists independently, For grouping, The number of phases in the system. The number of elements to consider for the system. The number of moles of the element. The atomic matrix of the components, The number of moles of the component. For components Standard enthalpy, For heat loss, For component indexing, For the first The number of moles of each component For the index of phase, For the first The first Partial molar free energy of the components For the first The first The number of moles of each component For the first The first The number of moles of each component For the first The number of moles of each component This refers to the standard enthalpy of formation of each component in the feed at 298K under standard conditions. For the first The components at their feed temperature The actual enthalpy below, For the first product The standard enthalpy of formation of the components at 298 K under standard conditions. For the first The components at the product temperature The actual enthalpy below.

[0010] Optionally, the phase balance includes: ; ; in, To minimize free energy This is to determine the existence conditions of each phase in the system, such as the liquid and gas phases. for The number of moles of an element For components medium elements The number of For components exist The number of moles in the phase, The total number of types of elements. The number of phases in the system. For the index of phase, This represents the total number of chemical components in the system.

[0011] Optionally, the objective function includes: ; in, For objective function 1, For objective function 2, For objective function 3, This represents the actual oxygen content in the flue gas. This is the simulated value for the oxygen content in the flue gas. This represents the actual CO2 content. The values ​​are from the CO2 content simulation model. This represents the actual H2O content in the flue gas. The values ​​are from the simulation model of H2O content. This is the temperature correction value for the RGibbs module.

[0012] Optionally, obtaining the optimal temperature setpoint includes: Step 1: Initialize the population; Step 2: Map the individuals in the population to a preset temperature range to obtain the target temperature value, and assign it to the Aspen Plus model to calculate fitness; Step 3: Perform adaptive selection, crossover, and mutation on the current population to update the population individuals; Step 4: Repeat steps 2-3 until the maximum number of iterations is reached. Extract the population individuals corresponding to the optimal fitness, decode the population individuals into temperatures, and use them as the optimal temperature setting value.

[0013] Optionally, the method further includes: Save the Aspen Plus model as a target format file, and when invoked, use the actxserver command to create an ActiveX object to drive the Aspen Plus model.

[0014] The beneficial effects of this invention are as follows: This invention introduces an adaptive genetic algorithm (AGA) to automatically optimize and correct key parameters in the MSWI full-process simulation model (such as the number of moles of components in chemical reaction equilibrium, phase distribution in phase equilibrium, heat loss coefficient, etc.) using actual field operation data. This allows the model parameters to be dynamically adjusted according to changes in operating conditions, effectively overcoming simulation deviations caused by fixed parameter settings. As a result, the error between the flue gas composition (O2, CO2, SO2, NO, etc.) and temperature distribution output by the model and the actual values ​​is significantly reduced under fluctuating operating conditions such as different waste compositions, moisture content, and calorific value, thereby improving the operating condition adaptability and prediction accuracy of the simulation model.

[0015] During the calibration process, the adaptive genetic algorithm of this invention can quantitatively evaluate the sensitivity and contribution of each parameter to be calibrated to the simulation output by constructing a fitness function and performing multi-generational evolution optimization. It realizes the identification and analysis of key influencing parameters, clarifies the influence of different physical property parameters, flow parameters and geometric parameters on the incineration effect, and provides data support for model simplification, operation optimization and on-site operation guidance.

[0016] This invention designs a unified interaction and data interface that connects MSWI numerical simulation software (such as Aspen Plus, Fluent, etc.) with adaptive genetic algorithm programs (such as MATLAB, Python), enabling automatic assignment of simulation model input parameters, automatic extraction of output results, and closed-loop operation of optimization iteration. It breaks down the interaction barriers between different software platforms, automates the parameter calibration process, eliminates the need for repeated manual parameter adjustments, and significantly improves calibration efficiency and engineering practicality.

[0017] This invention does not require derivative information of the objective function and is suitable for complex simulation models of MSWI processes with strong nonlinearity, multiphase equilibrium, and chemical reaction coupling. It has the advantages of strong global optimization ability and is not easy to get trapped in local optima, which significantly improves the engineering application value and credibility of the model. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of the MSWI full-process simulation model parameter correction method based on adaptive genetic algorithm according to an embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the interaction between MATLAB and Aspen Plus in an embodiment of the present invention; Figure 3 This is a flowchart of the model parameter optimization process based on Aspen Plus and MATLAB in an embodiment of the present invention; Figure 4 This is a schematic diagram of the variable optimization numerical process under grate speed according to an embodiment of the present invention; Figure 5 This is a schematic diagram of the variable optimization numerical process under the primary wind ratio in an embodiment of the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0022] like Figure 1As shown, this embodiment discloses a parameter correction method for the MSWI full-process simulation model based on an adaptive genetic algorithm, including: obtaining the MSWI full-process numerical simulation model, determining the chemical reaction equilibrium and phase equilibrium of the RGibbs module in the MSWI full-process numerical simulation model; performing parameter analysis on the chemical reaction equilibrium and phase equilibrium, determining the temperature setpoint as the correction parameter, and taking the minimization of gas at G3 in the MSWI process as the objective function, iteratively solving the correction parameter through an adaptive genetic algorithm to obtain the optimal temperature setpoint.

[0023] Specifically, this embodiment discloses a parameter correction method for the MSWI full-process simulation model based on an adaptive genetic algorithm, including: Error Analysis: In the MSWI full-process multi-software coupled simulation strategy, the temperature setpoint of the RGibbs module is set based on the average value of the FLUENT furnace temperature simulation. The combustion process is assumed to be carried out under isothermal and isobaric conditions. The RGibbs module is analyzed based on the principle of minimum Gibbs free energy. For complex chemical reaction systems, the Gibbs free energy of the system will reach its minimum value when the reaction reaches thermal equilibrium. The mechanism is as follows.

[0024] (1) Chemical Reaction Equilibrium: Chemical equilibrium is determined by considering input components, reaction thermodynamic data, and reaction rates to determine the mole fraction of various gaseous components in a system to satisfy the equilibrium conditions of a chemical reaction. For a complex chemical reaction system, after the reaction reaches thermal equilibrium, the Gibbs free energy of the system is... The minimum value is reached. This nonlinear mathematical programming problem is described by the following equation: (1); The constraints are as follows: Enthalpy equilibrium constraint: (2); Enthalpy equilibrium constraint: (3); in, A phase that exists independently, such as solid particles; The number of phases in the system. For grouping, The number of elements to consider for the system. The number of moles of the component. The atomic matrix of the components, The number of moles of the element. For components Standard enthalpy; For temperature; This is due to heat loss.

[0025] (2) Phase equilibrium: Phase equilibrium determines the existence conditions of each phase in a system, such as the liquid and gas phases, by minimizing the Gibbs free energy. The MSWI process is also a problem involving phase equilibrium calculations of a chemical reaction system. Its mathematical description is: (4); in, For grouping, For phase number, and They are respectively Components in The number of moles and chemical potential in the phase.

[0026] The phase equilibrium calculation problem can be transformed into a nonlinear programming problem with the following linear constraints: (5); (6); in, for The number of moles of an element Components medium elements The number of.

[0027] As can be seen from the above formula, numerous parameters are involved in the calculation process. For example, a single phase... Number of phases in the system Groups Number of system elements Number of moles of components Component Atom Matrix Number of moles of elements Components Standard enthalpy ;temperature Heat loss Groups Components exist Number of moles in phase and chemical potential RGibbs module pressure and RGibbs module temperature Aspen Plus's RGibbs module uses the principle of minimizing Gibbs free energy to calculate phase and chemical equilibria by considering input conditions and system composition. During the calculation, it adjusts the molar fraction of various components to achieve the lowest possible Gibbs free energy.

[0028] Therefore, all the above parameters affect the simulation results. In the proposed multi-software coupling strategy, the input to the RGibbs module comes from the solid-phase combustion section, with components including CO, NO, N2, CO2, H2O, H2, CH4, S, and HCN. The temperature, pressure, total flow rate, and mass fraction of the input components are known set values. The output gas components of the RGibbs module are specified as CO, CO2, H2O, NO, NO2, HCN, SO2, and SO3. The number of phases in the module is known. In process simulation software such as Aspen Plus, reaction thermodynamic data are usually provided as constant values, especially at standard temperatures and pressures. However, for non-standard conditions such as high temperature and high pressure, it may be necessary to consider the effects of temperature and pressure, using thermodynamic data models related to temperature and pressure. The molar number, chemical potential, and standard enthalpy of the gas components are calculated based on the system conditions and input data to ensure thermodynamic and chemical equilibrium are met. Therefore, during the simulation, the molar number, chemical potential, and standard enthalpy are usually calculated internally by the module and vary according to the system conditions. It is influenced by factors such as module temperature and pressure, initial composition, and reaction thermodynamics data. Reaction thermodynamics data includes a series of physical quantities related to thermodynamic properties and thermochemical reactions. These data are used to describe the energy changes and equilibrium conditions of chemical reactions. Key reaction thermodynamic data include: heat of reaction, enthalpy of reaction, entropy of reaction, and change in reaction free energy. These data are typically functions of temperature and pressure, and therefore their values ​​may vary under different temperature and pressure conditions.

[0029] Module temperature has a significant impact on chemical equilibrium. In the proposed multi-software coupled simulation strategy, the setpoint of the RGibbs module temperature is based on the average value of the FLUENT furnace temperature simulation. While this method effectively solves the problem of how to set the RGibbs module temperature, this value is not necessarily the optimal solution and may not optimally characterize the temperature environment of the combustion reaction. Based on the specified simplification assumptions, some parameters can be understood as constants, such as the single phase S in the system, the number of phases in the system, and the number of component types considered. These parameters are not corrected. Therefore, this invention optimizes and adjusts the temperature setpoint of the RGibbs module, using the temperature setpoint parameter as a correction parameter. The optimal temperature setpoint is then sought sequentially to enable the model to more accurately characterize the actual combustion conditions.

[0030] Correction Strategy: Waste incineration simulation models can obtain simulated incineration effect data by inputting incineration process data. If the input parameter values ​​of the simulation model are not set reasonably, there will be errors between the actual measured values ​​and simulated values ​​of G3 flue gas. For example... , and The quality score.

[0031] This invention transforms the parameter correction problem into an optimization problem. First, the parameter to be corrected is used as the optimization parameter, and then an optimization algorithm is employed to solve it. This invention selects the temperature setpoint of the RGibbs module as the correction parameter, thus implementing single-parameter correction. The objective function is to minimize the difference between the model simulation values ​​of the gas (e.g., O2, CO2, H2O) at point G3 and the actual output of the plant. RGibbs module temperature correction value The fluctuation range is [900, 1200]. The optimization objective and constraints are as follows: (7); in, This represents the actual oxygen content in the flue gas. This is the simulated value for the oxygen content in the flue gas. This represents the actual CO2 content. The values ​​are from the CO2 content simulation model. This represents the actual H2O content in the flue gas. The values ​​are from the simulation model of H2O content. This is the temperature correction value for the RGibbs module.

[0032] Software interaction: The interaction method between different software programs. Interaction between different software programs is necessary when modifying parameters. Software interaction strategies are required.

[0033] like Figure 2 As shown, an interface is needed to enable interaction between Aspen Plus and MATLAB. In this experiment, the AspenPlus file is saved in bkp format for easy MATLAB access. When accessing the file, the `actxserver` command is used to create an ActiveX object, which is then opened in MATLAB to retrieve the model built by Aspen.

[0034] like Figure 3 As shown, the optimization steps are as follows: 1) Initialize the population. 2) Initialize the Aspen Plus model, load the file in the specified path, set the AGA parameters, maximum number of iterations to 10, crossover rate to 0.8, mutation rate to 0.1, and temperature range to [900, 1200]. 3) Map the different generated individuals of the population to values ​​within the temperature range, assign these values ​​to Aspen Plus, and calculate the fitness. 4) Perform adaptive selection, crossover, and mutation on the current population to update the individuals. 5) Repeat steps 3) and 4) until the preset maximum number of iterations is reached. 6) Find the individual with the best fitness among all generations, decode it into temperature, and output it as the optimal value.

[0035] In one embodiment, the experimental data used in the simulation model comes from actual data collected by the DCS system of a waste incineration plant. The primary air temperature is 500K, and the air volume is 68,000 m³ / h; the secondary air temperature is 300K, and the air volume is 7,000 m³ / h. The excess air coefficient is 0.9. The ratio of grate air volume in the drying section, primary combustion section, secondary combustion section, and combustion section is 0.24:0.50:0.20:0.06. The incineration capacity is set to 87.75% of the rated capacity. The grate velocity is 8 m / h. The program runs using MATLAB R2023 and Aspen Plus V11, with a computer CPU configuration of a 12th generation Intel(R) Core(TM) i7-12700K 3.60 GHz, 32GB of RAM, and a 4TB hard drive. The maximum number of iterations for AGA is 10, the crossover rate is 0.8, the mutation rate is 0.1, the encoding bit length is 6, and the maximum population size is 10.

[0036] The correction error is calculated as follows: (8); in, It is a correction error. It is the temperature value being corrected. It is a temperature value.

[0037] The formula for calculating the model fitness rate is defined as follows: (9); in, For the fitness rate of the simulation model, The data number used to establish the parameter calibration model. For data numbering used to establish parameter calibration models, Represents the actual value. Represents the simulated value. This represents the total number of experimental data.

[0038] Based on the constructed MSWI full-process simulation benchmark model, experiments were conducted sequentially under different road sign speeds and air volume ratios, resulting in 8 sets of simulation experimental data, as shown in Table 1.

[0039] Table 1 Non-standard operating conditions Numerical simulation results: Simulation data under baseline operating conditions were obtained by running the simulation model. This invention also collected actual factory operating data. A comparison between the simulation data and the actual data is shown in Table 2.

[0040] Table 2 Comparison between numerical simulation and actual data Table 2 shows that the relative errors between the simulated and actual flue gas temperature values ​​calculated according to the formula are 0.2%, the relative errors for O2 concentration are 20.0%, and the relative errors for CO2 concentration are 53.8%. These data indicate that the accuracy of the model established in this invention is within an acceptable range and can simulate the waste incineration process to a certain extent.

[0041] Calibration results: Under the multi-software coupled simulation strategy, eight sets of tests were simulated under non-benchmark conditions to obtain experimental data and verify the calibration effect under various different working conditions.

[0042] Depend on Figure 4 It can be seen that as the grate speed increases, the temperature correction value first decreases and then increases. Among the four discrete points, the maximum change is 10.11%, the minimum is -7.55%, and the average is 3.06%. Figure 5 It can be seen that as the proportion of primary wind gradually increases, the temperature correction value shows a trend of first decreasing and then increasing. Among the four discrete points, the maximum change is 3.43%, the minimum is -16.86%, and the average is -8.44%.

[0043] Therefore, when using a coupled modeling strategy, the temperature setpoint needs to be appropriately decreased when the grate speed gradually increases from 6 m / h to 8 m / h, and appropriately increased when the grate speed gradually increases from 8 m / h to 12 m / h. Similarly, the temperature setpoint needs to be appropriately decreased when the primary air ratio gradually increases from 0.84 to 0.88, and appropriately increased when the primary air ratio gradually increases from 0.88 to 0.96, in order to improve the model's prediction accuracy.

[0044] Based on the simulation experiment of the calibration model, the changes in the content of gas components in G3 flue gas are shown in Table 3.

[0045] Table 3. Changes in simulated gas content values ​​in G3 flue gas As shown in Table 3, after correction, the change in O2 mass fraction was 0.0072%, and the change in mass flow rate was 8.34 kg / h; the change in CO2 mass fraction was 0.0008%; and the change in H2O mass fraction was 0.0082%, and the change in mass flow rate was 10.90 kg / h.

[0046] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made to the technical solutions of the present invention by those skilled in the art without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A parameter calibration method for the MSWI full-process simulation model based on adaptive genetic algorithm, characterized in that, include: Obtain the MSWI full-process numerical simulation model and determine the chemical reaction equilibrium and phase equilibrium of the RGibbs module in the MSWI full-process numerical simulation model; The chemical reaction equilibrium and the phase equilibrium are analyzed by parameters to determine the temperature setpoint as the correction parameter. The optimal temperature setpoint is obtained by iteratively solving the correction parameter using an adaptive genetic algorithm with the goal of minimizing the gas at G3 in the MSWI process.

2. The MSWI full-process simulation model parameter correction method based on adaptive genetic algorithm according to claim 1, characterized in that, The chemical reaction equilibrium includes: ； ； ； in, The free energy of the system after the reaction reaches thermal equilibrium. Reaching the minimum value, As a phase that exists independently, For grouping, The number of phases in the system. The number of elements to consider for the system. The number of moles of the element. The atomic matrix of the components, The number of moles of the component. For components Standard enthalpy, For heat loss, For component indexing, For the first The number of moles of each component For the index of phase, For the first The first Partial molar free energy of the components For the first The first The number of moles of each component For the first The first The number of moles of each component For the first The number of moles of each component This refers to the standard enthalpy of formation of each component in the feed at 298K under standard conditions. For the first The components at their feed temperature The actual enthalpy below, For the first product The standard enthalpy of formation of the components at 298 K under standard conditions. For the first The components at the product temperature The actual enthalpy below.

3. The MSWI full-process simulation model parameter correction method based on adaptive genetic algorithm according to claim 1, characterized in that, The phase equilibrium includes: ； ； in, To minimize free energy This is to determine the existence conditions of each phase in the system, such as the liquid and gas phases. for The number of moles of an element For components medium elements The number of For components exist The number of moles in the phase, The total number of types of elements. The number of phases in the system. For the index of phase, This represents the total number of chemical components in the system.

4. The MSWI full-process simulation model parameter correction method based on adaptive genetic algorithm according to claim 1, characterized in that, The objective function includes: ； in, For objective function 1, For objective function 2, For objective function 3, This represents the actual oxygen content in the flue gas. This is the simulated value for the oxygen content in the flue gas. This represents the actual CO2 content. The values ​​are from the CO2 content simulation model. This represents the actual H2O content in the flue gas. The values ​​are from the simulation model of H2O content. This is the temperature correction value for the RGibbs module.

5. The MSWI full-process simulation model parameter correction method based on adaptive genetic algorithm according to claim 1, characterized in that, Obtaining the optimal temperature setpoint includes: Step 1: Initialize the population; Step 2: Map the individuals in the population to a preset temperature range to obtain the target temperature value, and assign it to the Aspen Plus model to calculate fitness; Step 3: Perform adaptive selection, crossover, and mutation on the current population to update the population individuals; Step 4: Repeat steps 2-3 until the maximum number of iterations is reached. Extract the population individuals corresponding to the optimal fitness, decode the population individuals into temperatures, and use them as the optimal temperature setting value.

6. The MSWI full-process simulation model parameter correction method based on adaptive genetic algorithm according to claim 5, characterized in that, The method also includes: Save the Aspen Plus model as a target format file, and when invoked, use the actxserver command to create an ActiveX object to drive the Aspen Plus model.

Images