On Stress Control Of Steam Turbine Rotor

Steam turbine is one of the most important equipment in thermal power generation and nuclear power generation equipment.The performance of steam turbine and the system regulation of the whole steam turbine can play a vital role in the safe and economic operation of the whole unit.The rotor of steam turbine is one of the most important parts of the whole steam turbine unit.The life of the rotor of steam turbine is directly related to the life of the whole unit.The rotor of steam turbine is the core component of the operating components,so it is very necessary and important to study the rotor of steam turbine.In the past,the analysis of steam turbine rotor is carried out through simple theoretical model,or through two-dimensional finite element model.The error of these methods is relatively large.In this context,through the analysis and improvement of the previous model,a three-dimensional finite element analysis model which is very close to the actual rotor is established,making the analysis results more accurate to create conditions for the follow-up research and experimental process.The main research work of this paper is as follows:（1）Taking a domestic 300MW steam turbine as the research object,according to the operation guidelines of its actual working conditions,the cold start mathematical model of300MW steam turbine rotor with constraints is designed.After the subsequent use of the model,the usability of the mathematical model is clarified.The subsequent work is based on the model,combined with particle swarm optimization algorithm and support vector machine to analyze the model Finally,the model is verified to have high accuracy.（2）Based on the actual working state of a domestic 300MW steam turbine rotor,the three-dimensional model of the rotor is established by using ANSYS finite element analysis software,and the boundary conditions are set according to the physical properties of the rotor material.Finally,the model is verified and analyzed by the starting data of the actual working condition,and the temperature field and stress field of the rotor model are calculated respectively,and compared with the actual working condition Compared with the actual data,the results of the final finite element analysis model are very consistent with the actual results.It shows that the3D model has high accuracy and usability.（3）t₁、t₂、t₃、t₄、t₅、t₆ are used to represent the starting time of this stage,The optimization results are carried out under actual constraints.Fifteen cold starting conditions were established,and the rotor stress field was analyzed by ANSYS software.Based on the finite element analysis results of 15 starting schemes,the prediction models of starting time and maximum thermal stress on rotor surface were established by using BP neural network and two kinds of support vector machines.Two kinds of SVM models are Gaussian kernel and Laplace kernel.At the same time,three prediction models are compared.Finally,it is proved that the support vector machine based on Laplace kernel function has a better prediction effect,and the prediction error is less than 0.3%.On the basis of comprehensive consideration of the results of the prediction model,the particle swarm optimization algorithm is used to solve the initial parameters.Particle Swarm Optimization（PSO）adopts adaptive weighted particle swarm optimization and improves the results of traditional PSO.The comparison of the two results highlights the advantages of the adaptive weighted particle swarm optimization algorithm.（4）In order to study the stress control of steam turbine rotor,the starting time of steam turbine rotor was optimized,and the relationship between the starting time and starting rules and the stress of the rotor was found through the verification of 3-D finite element model.The temperature and stress of the rotor were calculated and analyzed by using ANSYS.Taking the data calculated by the above optimization scheme as the sample data,the BP neural network and support vector machine were used to establish a prediction model to predict the ultimate stress value of the maximum stress point.Taking the results of the prediction model into consideration,the starting parameters are recovered by using the particle swarm optimization algorithm.The experimental results show that the starting parameters are verified by ANSYS within the acceptable range of the maximum stress.In the process of optimizing the starting time of steam turbine,the starting time of steam turbine rotor stress is greatly shortened within a reasonable range,and the starting efficiency of the unit is improved.（5）Steam turbine rotor’s temperature field and stress field modeling,system identification method is used to fitting the transfer function of input and output data model,at the same time,the transfer function model discretization,using input data verification,validation after the model is reasonable,the model into the form of state space model,using the MPC model predictive control algorithm of steam turbine rotor stress control problem is studied.First,the temperature model was identified with an accuracy of 99.73%,and then the stress model was identified with an accuracy of 98.2%.Then discretize the transfer function,and input the data into the discretized transfer function to verify the accuracy of system identification.After the accurate verification of the transfer function,the transfer function was transformed into the dynamic matrix control model,and the stress feedback controller was designed to realize the stress predictive control（MPC）of the rotor model.At the same time,the output responses of the system and the controller before and after the addition of the controller were compared respectively.Finally,the rationality of the control model is verified.The three-dimensional finite element model established in this paper greatly improves the simulation accuracy.Combined with the mathematical model,the function model of temperature field and stress field,and the model predictive control algorithm,the stress control problem of steam turbine rotor is studied.Under the new stress control idea,the final control effect is better than the traditional control effect.