针对实施鲁棒参数控制的过程模型中存在的观测不确定性和模型不确定性,提出了一种谨慎控制律的设计方法.通过概率统计计算,将噪声因素观测误差和回归模型系数估计误差引入控制模型,明确考虑在误差对系统影响的情况下如何对过程参数进行控制.谨慎控制律会根据误差大小确定参数的在线调节幅度,使系统性能在不确定性环境中表现得更一致和健壮.拓展了谨慎控制理论的应用范围,改善了鲁棒参数控制方法在过程存在不确定因素情况下的性能.实例分析表明,上述方法是有效且实用的.
In order to take the observation uncertainty and model coefficient uncertainty into consideration, an integrated design of cautious control with robust parameter control is proposed in this article. Through statistical calculation, observation errors of noise factors and estimation errors of regression model parameters are introduced into the control model. Cautious control, which explicitly considers these two kinds of uncertainty in adjusting the online controllable factors, makes the system performance consistent and stable in such uncertain conditions by lessening control actions if the errors are large. The proposed approach enlarges the application of cautious control theory and improves the performance of robust parameter control in robustness to the uncertainty. A case study is presented to verify that the proposed cautious control is effective and practical.