Search

CN-122021241-A - Method for analyzing and optimizing decision-making cause of flue gas flow deviation of electric precipitation inlet

CN122021241ACN 122021241 ACN122021241 ACN 122021241ACN-122021241-A

Abstract

The invention relates to the technical field of flue gas treatment equipment and discloses a method for analyzing and optimizing and deciding the flue gas flow deviation cause of an electric dust removal inlet, which comprises the following steps of establishing a three-dimensional flow field calculation model and defining a continuous operation working condition space; generating a parameterized model of the flow guiding device driven by design variables, constructing a response surface agent model for predicting flow deviation and total pressure drop according to working conditions and the design variables, performing multi-objective robust optimization based on the agent model to obtain a Pareto optimal front, selecting a final design scheme according to the Pareto optimal front, acquiring actual industrial KPI and flow deviation in real time after operation, calculating model prediction residual errors, and updating the response surface agent model on line by using measured data when the prediction residual errors exceed a threshold value. According to the invention, through multi-objective robust optimization and online self-adaptive correction, the high performance and full life cycle self-adaptive capacity of the flow guiding device under the full working condition are realized, and the design robustness is improved.

Inventors

  • MA QINLIANG
  • ZHANG YUEMING
  • XIAO FUZHEN

Assignees

  • 华电国际电力股份有限公司邹县发电厂

Dates

Publication Date
20260512
Application Date
20251219

Claims (10)

  1. 1. The method for analyzing and optimizing the cause of the flue gas flow deviation at the electric precipitation inlet is characterized by comprising the following steps: s1, establishing a three-dimensional flow field calculation model of an inlet flue of an electric dust collector, and defining a continuous operation condition parameter space; s2, generating a parameterized model of the flow guiding device driven by a set of design variables; S3, constructing a response surface agent model, wherein the response surface agent model is used for predicting a flow deviation coefficient and the total pressure drop of the system according to the operation condition parameter space and the design variable; S4, solving a multi-objective robust optimization problem based on the response surface agent model to determine an optimal solution set of the design variables, thereby obtaining a Pareto optimal front; s5, selecting a final design scheme of the flow guiding device according to the Pareto optimal front edge; S6, after the diversion device is put into operation, acquiring actual operation condition parameters and actual measurement flow deviation coefficients in real time, and calculating prediction residual errors between the actual measurement flow deviation coefficients and the response surface agent model prediction values; and S7, when the prediction residual continuously exceeds a preset threshold, the collected data comprising the actual operation condition parameters and the actual measurement flow deviation coefficients are utilized to update the response surface agent model on line.
  2. 2. The method for analyzing and optimizing the decision-making of the flue gas flow deviation at the inlet of the electric precipitation according to claim 1, wherein the step of generating the parameterized model of the deflector comprises: Defining a design domain in the three-dimensional flow field calculation model, and generating an initial configuration of the flow guiding device by adopting a topology optimization algorithm; Based on the initial configuration, key geometric features are extracted and correlated with the design variables to build the parameterized model.
  3. 3. The method for analyzing and optimizing the decision-making of the flue gas flow deviation of the electric precipitation inlet according to claim 1, wherein the response surface agent model is a kriging model; the method comprises the steps of generating sample point combinations in a joint parameter space formed by an operation working condition parameter space and a design variable space formed by the design variables by adopting an optimal Latin hypercube sampling method, obtaining flow deviation coefficients and total pressure drop of a system corresponding to each sample point combination by utilizing calculation fluid dynamics batch calculation, and generating the Kriging model by utilizing calculation results training.
  4. 4. The method for analyzing and optimizing the decision-making of the flue gas flow deviation cause of the electric precipitation inlet according to claim 1, wherein the optimizing objective of the multi-objective robust optimizing problem comprises: minimizing mathematical expectations of the flow deviation coefficient within the operating condition parameter space; minimizing a mathematical variance of the flow deviation coefficient within the operating condition parameter space; A mathematical expectation that minimizes the total pressure drop of the system within the operating condition parameter space.
  5. 5. The method for analyzing and optimizing the decision-making of the flue gas flow deviation of the electric precipitation inlet according to claim 4, wherein said step of solving a multi-objective robust optimization problem comprises: And solving the multi-objective robust optimization problem by adopting a non-dominant ordering genetic algorithm II to obtain the Pareto optimal front.
  6. 6. The method for analyzing and optimizing the decision-making of the flue gas flow deviation at the inlet of the electric precipitation according to claim 1, wherein the step of selecting the final design of the deflector comprises: Respectively assigning weight coefficients to each optimization target of the multi-target robust optimization problem, and constructing a unified utility function; And comprehensively scoring all solutions on the Pareto optimal front by using the utility function, and selecting a solution with an optimal utility value as the final design scheme of the flow guiding device.
  7. 7. The method for analyzing and optimizing the cause of the deviation of the flue gas flow at the inlet of the electric dust collector according to claim 1, wherein the flow deviation coefficient is calculated by calculating the standard deviation of the flue gas flow at all the inlet channels of the electric dust collector and dividing the standard deviation by the average flow of the flue gas flow at all the inlet channels.
  8. 8. The method for analyzing and optimizing the decision-making of the flue gas flow deviation at the inlet of the electric precipitation according to claim 1, wherein the step of online updating comprises the steps of: Taking the data points comprising the actual operation condition parameters and the actually measured flow deviation coefficients as new training samples, and merging the new training samples into an original sample data set to form an enhanced data set; Retraining the response surface proxy model using the enhanced data set.
  9. 9. The method for analyzing and optimizing and deciding the cause of the flue gas flow deviation at the inlet of the electric precipitation according to claim 1, further comprising the step of S8 model adaptive correction after the response surface agent model is updated on line, wherein the model adaptive correction comprises the following steps: And providing dynamic decision support based on the updated response surface agent model, wherein the dynamic decision support comprises predictive early warning according to future predicted working conditions or performance benchmark calibration according to current real-time working conditions.
  10. 10. The method for analyzing and optimizing the decision-making of the flue gas flow deviation cause of the electric precipitation inlet according to claim 2, wherein the optimization objective of the topology optimization algorithm is a weighted combination of flow distribution uniformity and flow pressure drop.

Description

Method for analyzing and optimizing decision-making cause of flue gas flow deviation of electric precipitation inlet Technical Field The invention relates to the technical field of flue gas treatment equipment, in particular to a method for analyzing and optimizing decision-making of flue gas flow deviation of an electric precipitation inlet. Background In the industrial fields of thermal power generation and the like, an electric dust collector is core equipment for purifying flue gas and controlling particulate emission. The operation efficiency is directly related to the uniformity of the velocity distribution of the inlet flue gas, and obvious flow deviation can cause the problems of overhigh flow velocity of the flue gas in a local area, secondary dust emission, uneven electric field distribution and the like, thereby severely restricting the overall dust removal performance. Therefore, an effective flow guiding device is designed and additionally arranged in an inlet flue of the electric dust collector so as to optimize the flowing state of the smoke, and the flow guiding device becomes a key technical measure for guaranteeing the high-efficiency operation of the electric dust collector and meeting the increasingly strict environmental protection standard. Currently, the design of an electrostatic precipitator inlet flow guide device relies primarily on engineering experience in combination with Computational Fluid Dynamics (CFD) simulation. The conventional flow is that an engineer firstly establishes a three-dimensional model of an inlet flue according to a design drawing, manually presets the initial layout and the form of a guide plate in the model by virtue of professional knowledge and hydrodynamic understanding, then carries out numerical simulation on a plurality of preset and representative unit operation working conditions (such as 100% rated load and 75% load) by utilizing CFD software, evaluates the effectiveness of a design scheme by analyzing simulation results, and carries out repeated manual adjustment and recalculation according to the effectiveness until a scheme meeting the design requirement is obtained. Although the prior art can improve the flow distribution of the inlet to a certain extent, the prior art has the defects of firstly, insufficient design robustness and full-working-condition adaptability. This is because the basis for design optimization is essentially a few discrete, static operating points, while the actual load of the genset is continuously and dynamically fluctuating over a wide range. The great computational cost of CFD simulation makes it impossible to perform traversal calculation on all possible working conditions in practice, which results in a scheme optimized for rated working conditions, and the flow conductivity of the CFD is possibly greatly compromised when the unit runs under low load or variable load due to the fundamental change of flow field characteristics. Secondly, the design flow is highly dependent on subjective experience of engineers, and is a repeated manual error testing process, so that the development period is long, the efficiency is low, more importantly, the advantages and disadvantages of the design scheme are limited in the existing cognitive framework of the engineers, and the whole design space is difficult to systematically explore to find out the breakthrough of the conventional global optimal configuration. Most importantly, the existing design is a one-time static scheme and cannot adapt to the physical state change of the equipment in the whole operation life cycle. The initial design is based on an idealized clean flue model, but in long-term operation, phenomena such as dust accumulation, abrasion, blockage and the like can continuously change the geometric form and flow resistance characteristics in the flue, and the prior art framework lacks an online monitoring feedback and model correction mechanism, cannot sense the performance degradation and provides self-adaptive adjustment, so that the difference between the actual operation effect and the initial design expectation is gradually enlarged with the lapse of time. Disclosure of Invention Aiming at the defects of the prior art, the invention provides an analysis and optimization decision method for the flue gas flow deviation of an electric dust removal inlet, which solves the problems that the existing flow guiding device is only designed for discrete working conditions, depends on subjective experience and cannot be subjected to online self-adaptive correction, and the robustness of the whole working conditions is poor and the long-term performance is declined. In order to achieve the purpose, the invention is realized by the following technical scheme that the method for analyzing and optimizing the cause of the flue gas flow deviation of the electric precipitation inlet comprises the following steps: s1, establishing a three-dimensional flow field calculation mod