On regularized feedback update of distributed-parameter systems in control and nonlinear estimation applications

This paper considers control of discretized linear distributed-parameter no-memory systems. It demonstrates that there is a need to regularize the optimality criterion in a feedback update design. It is shown that robustness of the update to a modelling error is proportional to the square root of the regularization parameter. Three applied control problems are discussed: 1) cross-directional control of the web manufacturing process; 2) iterative learning control of a batch process or repetitive motion; and 3) estimation of a nonlinear function of a single argument by using a radial basis function network.