基于改进和声搜索算法的主蒸汽温度现场数据辨识

Identification of Main Steam Temperature System Based on Improved Harmony Search Algorithm and Field Data

针对主蒸汽温度系统现场数据的模型辨识问题,提出了结合粒子群优化算法的改进和声搜索算法.采用经验模态分解法对带噪声污染的现场数据进行滤波处理,采用离散相似法进行模型辨识的计算机仿真实现和数值计算.应用该改进算法对循环流化床主蒸汽温度系统模型进行了现场数据辨识.结果表明:所辨识的模型具有较高的精度,能够反映实际主蒸汽温度系统的动静态特性;改进和声搜索算法比粒子群优化算法具有更好的稳定性和全局寻优能力,以及更快的收敛速度.

For model identification of main steam temperature system with field data,an improved harmony search(HS)algorithm was proposed in combination with particle swarm optimization(PSO).Empirical mode decomposition was used for filtering of noised data.Discrete-simularity method was chosen for computer implementation and numerical calculation after analyzing and comparing the commonly used identification strategies.The proposed method was finally applied for model identification of the main steam temperature system in a circulating fluidized bed.Results show that the obtained model has a high indentification precision,which can reflect the dynamic and static characteristics of actual main steam temperature systems;with fast convergence speed,the improved HS algorithm exhibits better stablility and higher global optimization capability,compared with PSO method.