不确定控制系统概率鲁棒性分析——自适应重要抽样法  

Probabilistic robustness analysis of uncertain control systems using adaptive importance sampling

作　　者：吴淮宁[1] 蔡开元[1] 

机构地区：[1]北京航空航天大学自动化科学与电气工程学院,北京100083

出　　处：《控制理论与应用》2004年第5期812-816,共5页Control Theory & Applications

基　　金：国家自然科学(青年)基金资助项目(60204011);国家自然科学基金资助项目(60274057).

摘　　要：将自适应重要抽样(AIS)法应用于不确定控制系统的概率鲁棒性分析问题,克服了标准MonteCarlo仿真(MCS)方法不能有效解决小概率事件的困难.给出了一种新的AIS方案.首先采用了一种递归的估计条件众数算法来产生一组使得系统不稳定或性能不可接受的不确定参数向量样本.然后利用这组样本来估计初始高斯型重要抽样密度函数的参数,并执行随后的迭代仿真过程.仿真结果验证了该方法的有效性.Adaptive importance sampling (AIS) method is applied to probabilistic robustness analysis problem of uncertain control systems,in order to overcome the difficulty that the standard Monte Carlo simulation (MCS) method cannot efficiently deal with rare events.A new AIS scheme is presented.First,a recursive algorithm estimating conditional mode was employed to generate a set of uncertain parameter vector samples which lead to instability or unacceptable performance of systems.And then,the subsequent iterative simulation procedures were taken with initial Gaussian importance sampling density function whose parameters were estimated by using this set of samples.Simulation results were provided to verify the effectiveness of the proposed method.

关 键 词：不确定控制系统 鲁棒性分析 概率方法 重要抽样 MONTE Carlo仿真 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]