Iterative correlation-based controller tuning

摘要: This paper gives an overview on the theoretical results of recently developed algorithms for iterative controller tuning based correlation approach. The basic idea is to decorrelate output error between achieved and designed closed-loop systems by iteratively parameters. Two different approaches are investigated. In first one, a equation involving vector instrumental variables solved using stochastic approximation method. It shown that, with appropriate choice finite number data at each iteration, algorithm converges solution equation. convergence conditions derived accuracy estimates studied. second approach minimization criterion. frequency analysis criterion shows that two norm desired transfer functions minimized independently noise characteristics. leads definition generalized which allows mixed sensitivity problem be handled in norm.