一类随机稳态优化算法的鲁棒性

ROBUSTNESS OF STOCHASTIC OPTIMIZATION ALGORITHMS FOR STEADY-STATE INDUSTRIAL PROCESSES

给出了随机稳态优化算法的灵敏度分析，讨论算法解对某些参数变化的依赖关系．文中给出了仿真例子。

In order to use the dynamic-response information for steady-state optimizing control ofindustrial processes,a new algorithm for solving the stochastic steady-state optimizing con-trol was presented in recent literature, The algorithm has reduced the required controller setpoint change in each on-line iteration and much less sensitive to noise, This paper gives anapproach to investigate the sensitivity of the algorithm, and discusses the dependence of theoptimal solution from the algorithm on the parameters. The robust performance of two algo-rithms is also compared. SimulatiOn results show the conclusion in some sense.