基于综合算法的汽轮机及其调节系统参数智能辨识

Intelligent parameter identification of steam turbine and its governing systems based on multi-algorithm

基于Matlab/Simulink,开发了综合多项式拟合、最小二乘法、单纯形法、遗传算法、模拟退火算法的汽轮机及其调节系统参数智能辨识软件。软件能自动完成原始数据的预处理、执行机构参数辨识样本的提取、负载试验数据边界条件的处理等,继而采用多种算法分别对参数进行辨识,并根据预设的逻辑选出最终的参数辨识结果。仿真校核的结果表明,所得到的参数辨识结果能够反映实际汽轮机及其调节系统的综合特性。最后,指出了影响汽轮机及其调节系统参数辨识准确性的关键因素。

Based on Matlab/Simulink, the software for intelligent parameter identification of steam turbine and its governing system is developed, in which polynomial fitting, least square method, the simplex method, genetic algorithm, simulated annealing algorithm are applied. The pretreatment for the original data, intelligent extraction of actuator parameters identification sample, and intelligent processing of boundary conditions of load test data are performed automatically with this software. Furthermore, this software uses several algorithms to identity parameters, and selects the final results according to the preset logic. The simulation and verification results show that the parameters identified by this software can reflect the characteristics of actual steam turbine and its governing system. Finally, the key factors which influence the accuracy of the parameter identification are also indicated.