Gardner, B. E.; Cannon, R. H. jun.; Bryson, Arthur E. jun. Time-invariant controllers for target tracking. (English) Zbl 0623.93078 J. Guid. Control Dyn. 10, 330-337 (1987). A method is described for designing time-invariant controllers to minimize root-mean-square target-tracking errors with root-mean-square control constraints in the presence of random initial conditions, target motions, and sensor errors. The maximum control responses usually occur during the initial transient period before the statistical steady state is reached. Hence we determine the controller parameters by minimizing a quadratic performance index that has separate weights on the transient and steady-state components of the control and tracking error. The designer makes tradeoffs between maximum control responses and steady- state tracking errors by varying the weights in this performance index. Two examples are given. The first is relatively simple to demonstrate the ideas,while the second is more complex, involving the recovery of a remotely piloted vehicle in a net on a rolling, pitching ship. The controller designs exhibit an improved performance over designs based on the minimization of the conventional performance index associated with linear quadratic Gaussian and parameter optimization methods. MSC: 93E20 Optimal stochastic control 93C05 Linear systems in control theory 93B50 Synthesis problems 93C95 Application models in control theory Keywords:time-invariant controllers; target-tracking; random initial conditions, target motions, and sensor errors; quadratic performance index; maximum control responses; steady-state tracking errors; time-invariant PDFBibTeX XMLCite \textit{B. E. Gardner} et al., J. Guid. Control Dyn. 10, 330--337 (1987; Zbl 0623.93078) Full Text: DOI