基于改进遗传算法的单元机组非线性模型参数辨识

Parameters identification of thermal power unit plant nonlinear model based on improved genetic algorithm

针对火电厂单元机组的特点及遗传算法工具箱在辨识多变量、非线性系统参数中存在的早熟、收敛速度慢等问题,对遗传算法工具箱进行了改进,以单元机组非线性动态模型为研究对象,提出了基于改进遗传算法工具箱的参数辨识方法.根据托电600 MW机组的阶跃扰动试验数据,辨识得到了单元机组非线性动态模型的参数.结果表明改进遗传算法工具箱对单元机组非线性模型参数辨识具有良好的适应性,辨识得到的模型是有效可靠的.

Aiming at characters of the fossil-fired power plant unit and the precocious, and scalability problems of the genetic algorithm toolbox in identifying muhivariable nonlinear system parameters, the genetic algorithm toolbox was improved, the nonlinear dynamic model was chosen as the study subject, and the method of parameters identification based on improved genetic algorithm toolbox was pro- posed. According to the step disturbance test data of 600 MW Unit in Tuoketuo No. 2 Power Plant, the unit model parameters were identifiod. The results show that the improved genetic algorithm tool- box also has a good adaptability to identify parameters for the unit model, and the identified model is valid and reliable.