改进量子粒子群算法及其在系统辨识中的应用  

An Improved Quantum Particle Swarm Optimization and Its Application in System Identification

作　　者：黄宇[1] 韩璞[1] 刘长良[1] 李永玲[1] 

机构地区：[1]华北电力大学自动化系,河北省保定市071003

出　　处：《中国电机工程学报》2011年第20期114-120,共7页Proceedings of the CSEE

基　　金：华北电力大学青年基金项目(200811013)~~

摘　　要：针对量子粒子群算法(quantum particle swarm optimization,QPSO)的收敛速度和寻优精度问题,提出了一种改进的QPSO算法。采用混沌序列初始化量子的初始角位置;在算法中加入变异处理,有效地增加了种群的多样性,避免早熟收敛。函数优化测试结果表明:该文提出的算法具有良好的优化效果。同时利用该文提出的算法对经典的具有无限冲激响应(infinite impulse response,IIR)的自适应递归滤波器模型进行了辨识,辨识结果证明了这种算法的有效性。利用此算法,在结合某分散控制系统的基础上,编制出了一种通用的热工对象模型辨识算法模块,并应用于某循环流化床电厂的辨识,取得了令人满意的辨识结果。In order to improve convergence speed and precision of optimization in quantum particle swarm optimization （QPSO）, an improved QPSO algorithm was presented. Chaotic sequences were used to initialize the origin angle position of particle; Mutation algorithm was introduced, which can effectively increase diversity of population, and also can avoid premature convergence. The test results of function optimization show that the proposed algorithm has better optimized effect. The improved algorithm proposed was applied to identify the classic adaptive infinite impulse response （IIR） model, and results proved the validity of the algorithm. On the basis of distributed control system （DCS）, a general-purpose identification algorithm modular for thermal object model was programmed, and it was applied to the identification of circulating fluidized bed power plant, achieving satisfactory results.

关 键 词：量子粒子群算法 系统辨识 循环流化床锅炉 自适应无限冲激响应滤波器 分散控制系统 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]