Research On Prediction And Control Method Of Assembly Quality Of Steam Turbine Diaphragm

With the "Made in China 2025" strategy put forward,my country’s manufacturing industry is transforming and upgrading in the direction of informatization,digitization,and intelligence.Under this background,as a heavy-duty defense equipment,the demand for the transformation of the production model to intelligentization has become increasingly prominent.As a key part of the steam turbine,the diaphragm has a complicated assembly process and many factors that affect its assembly quality,making its production data utilization difficult and low utilization rate,which leads to the lack of quantitative basis for prior control of the diaphragm production process.It needs to be repaired repeatedly to achieve Quality requirements,low assembly efficiency.Data is one of the core elements of intelligent manufacturing,and its value creation is the key to giving manufacturing "intelligence".In order to achieve data-driven diaphragm production and ultimately achieve the goal of intelligent manufacturing,it is necessary to use intelligent methods to predict and control the assembly quality of steam turbine diaphragms and create value data for diaphragm quality.In this paper,the assembly of a diaphragm in a steam turbine plant is taken as the research object,and the following research is carried out to achieve effective prediction and control of the assembly quality of the turbine diaphragm.First,analyze the research status of key quality characteristics identification,assembly quality prediction,and assembly quality control at home and abroad,and establish a general research framework for the assembly quality prediction and control of the steam turbine diaphragm plate;through the assembly process and quality characteristics of the diaphragm plate Study to determine the characteristics and prediction goals of the diaphragm assembly quality data set;based on the characteristics of the diaphragm quality data set,compare and analyze the data mining prediction algorithm and the prediction model parameter optimization method to determine the construction of the diaphragm assembly quality with PSO-SVM Forecasting model.Secondly,in view of the high dimensional characteristics of the diaphragm data set,the feature selection method is used to screen the key quali ty characteristics that affect the quality of the diaphragm to achieve the dimensionality reduction effect and improve the calculation efficiency of the prediction model;through the study of the feature selection method,based on the quality of the diaphragm The data set presents a model for identifying key quality characteristics of diaphragms based on the Relief and GA-Wrapper chain combination algorithm.Thirdly,based on the partition quality data set,the SVM optimization problem with kernel function and soft interval is constructed;the PSO parameter optimization algorithm is used to optimize the SVM related parameters;based on the PSO-SVM,the partition assembly quality prediction model is constructed,and a prediction-based The model’s assembly quality control mode.Finally,using PyCharm,MySQL,DBeaver and other development tools,based on the Python programming language,design and develop a steam turbine diaphragm assembly quality prediction and control system,and take a steam turbine plant as an example of a unit diaphragm assembly for application verification.Through the key quality characteristic identification model of the diaphragm,analyze the key quality characteristics that affect the assembly quality of the diaphragm;on this basis,the data mining prediction algorithm is used to construct the assembly quality prediction model,which provides a quantitative basis for the advance control of the diaphragm assembly quality,and realizes the data drive Production,create data value,have a certain application value for the transformation and upgrading of the production of the diaphragm,and the realization of the goal of "smart manufacturing".