Simple autopilot design procedure based on a factorization approach

Abstract

PurposeThe purpose of this paper is to apply a factorization‐based control system design procedure to design of auto‐pilot systems.Design/methodology/approachThe design approach is based on a previously developed factorization‐based control system design procedure. The design approach requires a stable coprime factorization of the plant, a set of stable coprime factors that are solutions to the Bezout identity, the linear model of a desired response, and the desired frequency range of interest. When all this information is provided, the developed automated software outputs a candidate auto‐pilot whose performance should be verified.FindingsFor a specific class of aerospace vehicles, it is found that the described factorization‐based auto‐pilot design procedure may replace the presently complicated auto‐pilot design procedures. The final design is simpler than other techniques that are based on classical, robust, adaptive, QFT, gain scheduling, or interpolation techniques, and the total required man hours for the design loop is less than the mentioned alternative approaches.Research limitations/implicationsThere are two basic limitations – time variations of plant parameters must not be very large as is the case with the agile aerospace vehicles, and specification of the desired closed‐loop system behavior is not systematic.Practical implicationsThe major outcome of this research is that complicated autopilots of a class of aerospace vehicles may be replaced by simpler systems with competitive performance.Originality/valueThis is the first paper in the area of autopilot design that is based on the application of a simple factorization‐based design procedure.