基于鲁棒优化的系统辨识算法研究

For identification of system in which the data is uncertain and only known to belong to a bounded set instead of some probability distributions, a new method based on robust optimization is proposed. This method transforms system identification into uncertain optimization problem, then selecting appropriate uncertainty set, deriving and solving the computationally tractable robust counterpart, finally realizing identification of uncertain system. Simulation results show the feasibility and effectiveness of this method.