摘要
数字随动(伺服)系统因具有非线性、不确定性和时变性,难以精确地建立系统的数学模型,由此对设计高性能的控制器造成困难。遗传算法因其隐含并行性、全局搜索等优点,在优化问题领域得到了很好的应用,但遗传算法简单应用于如数字随动系统这样的控制系统在线闭环辨识或参数优化时,会面临大量重复实验成本高、实验时间过长引起系统不稳定等实际问题。针对以上困难,提出一种基于嵌入式仿真的系统辨识方法,采用遗传算法进行数字随动系统同步在线闭环模型参数辨识与控制参数寻优.经过在数字随动系统实物环境下实验,证明了该方法的有效性.
It is difficult to accurately establish the dynamic model of digital servo systems with nonlinearity, uncertainty, and time-varying, and as a result, a high-performance controller can not be designed easily. Because of advantages such as nature parallelism and global search, genetic algorithm was employed in solving optimization problems widely. Nevertheless, while genetic algorithm is simply used to closed-loop identification and parameters optimization of complex control system online, as a digital servo system, the practice problems will be faced, that cost of numerous repetitive experiment is high and experiment long will cause instability of system. To solve these problems, a method of embedded simulation for synchronous closed-loop ident^ication of model parameters and control parameters optimization of a digital servo system online was given based on genetic algorithm. Through the experiments under the environment of real-world apparatus of the digital servo system, the effective of this method was confirmed.
出处
《系统仿真学报》
CAS
CSCD
北大核心
2009年第21期6945-6949,共5页
Journal of System Simulation
基金
国家自然科学基金(60474039)
关键词
数字随动系统
遗传算法
嵌入式仿真
在线闭环辫识
参数寻优
digital servo system
genetic algorithm
embedded simulation
closed-loop identification
parameters optimization
