基于粒子群优化算法的过程模型辨识

Parameter identification of process model based on PSO

参数辨识是过程建模的基础,提出了一种基于粒子群优化(PSO)算法的模型参数辨识方法,将过程模型的每个参数作为粒子群体中的一个粒子,利用粒子群体在参数空间进行高效并行的搜索来获得过程模型的最佳参数值,可有效提高参数辨识的精度和效率。对火电厂热工过程进行参数辨识的仿真结果表明,利用PSO算法辨识过程模型参数,无论过程模型是否是时滞对象,该辨识方法对过程模型的阶次不敏感,对于不同的输入信号,均能得到满意的辨识精度和效率,因此得到了较为精确的过程模型,模型输出与实际输出基本一致。

Parameter identification is the base of process modeling,a PSO（Particle Swarm Optimization ） -based model parameter identification method is put forward. By taking every parameter of process model as particle of the swarm and applying PSO algorithm to search optimal parameters of the process model concurrently and efficiently in the parameter space,the precision and efficiency of parameter identification are improved effectively. Simulation results for power plant thermal process indicate that, for model with or without time lag,the method is not sensitive to the order of the model,and obtains satisfactory identification precision and efficiency for different input signals. The precise process model is thus built and model outputs coincide with actual outputs.