基于粒子群算法的汽轮机及其调速系统参数辨识方法

Parameter Identification Method of Steam Turbine and Its Speed Governor System Based on Particle Swarm Optimization

为了解决目前汽轮机及其调速系统参数辨识传统方法过程复杂、耗时长、人为干预较多等问题,提出了一种基于粒子群算法的一键式智能辨识方法。综合考虑机组实际运行参数偏离设计值等问题,根据现场实测数据,自动搜寻扰动特性参数及对应的响应,并采用自动整理后的参数作为粒子群算法的输入,完成一键式参数辨识过程。该方法既减少了人的劳动强度,也避免了人为误差的引入。将该辨识方法用于实际火电机组汽轮机及其调速系统的参数辨识,并与传统的粒子群算法辨识结果进行对比,表明了该辨识方法的实用性、方便性和高效性。本辨识方法为汽轮机调速系统的参数辨识提供了一种新的有效手段。

For reduction of the complexity of processes, time-consuming and manual labor in traditional parameter identification methods of steam turbine and its speed governor system, a new one-touch intelligent method based on particle swarm optimization was proposed. In consideration of the deviation between the actual operating parameters and the designed ones, the dynamic model was built. The practical data were used to get the efficient disturbance parameters and their responses automatically, then taking these selected data as input data, based on a modified particle swarm optimization scheme, the whole parameter identification process was completed in an one-touch way. This method could not only reduce human workload, but also avoid the importing human error. It was validated by using the method to complete the parameter identification of a specific steam turbine system and comparing with the traditional particle swarm optimization method. The results show that this method is more practical, convenient and efficient. It provides a new way for parameter identification of steam turbine and its speed governing system.