非线性系统参数估计的一类有效搜索策略

A Class of Effective Search Strategies for Parameter Estimation of Nonlinear Systems

结合模拟退火的随机概率突跳性搜索和单纯形法的凸多面体几何搜索 ,提出了非线性系统参数估计的一类有效搜索策略 .通过对多种非线性定常系统的多维参数估计和非线性时变水箱系统的参数与时滞在线联合估计的仿真研究 ,验证了该方法的可行性。

Combining the probabilistic jumping search of simulated annealing with the geometry search in convex polyhedron of simplex method, this paper proposes a class of effective search strategies for parameter estimation of nonlinear systems. Simulation results, including multiple parameter estimation of several nonlinear constant systems and joint parameters and time-delay on-line estimation of time-varying nonlinear water tank system, demonstrate the feasibility, effectiveness and certain anti-noise ability of the proposed strategies.