We consider a class of linear systems in which there is time-varying uncertainty. These linear uncertain systems can be divided into two types. Systems in which the structure of the uncertainty satisfies certain matching conditions are called matched, and those systems in which the uncertainty does not satisfy the matching conditions are called mismatched. A linear control law is determined which produces tracking of dynamic inputs. The tracking error does not asymptotically decrease to zero because the systems are uncertain, instead the error is bounded. In the case of matched systems this error bound can be made arbitrarily small, and the system is said to practically track the input. In mismatched systems, the tracking error cannot be made arbitrarily small, and the system is said to ε-track the input. Previously published theory requires nonlinear controllers for practical tracking. Here, we derive a linear feedforward control law. Several examples illustrate the results.

We consider a class of linear systems in which there is time-varying uncertainty and additive disturbances. A control law is determined which produces practical tracking. Practical tracking guarantees that the asymptotic tracking error can be made arbitrarily small.Existing theory provides asymptotic tracking only for systems where the uncertainty is unknown but constant. The theory presented here is able to accommodate the more realistic situation of timevarying uncertainty.

A standard electro-optical sensor can perform several different surveillance missions to support tactical military users. The missions include environmental sensing, land and ocean remote sensing, tactical missile tracking, and space object surveillance. The key is that while the spacecraft is a standard configuration for all missions, its design is a compromise between the specific requirements for each mission; the orbit chosen and operations mode for each mission also vary. Although sub-optimal for any given mission, standard sensor systems have the advantage of achieving a higher benefit-to-cost ratio by realizing economies of scale in production and reduced development. Point designs of three different multi-mission sensors are presented, supported by design analysis, and encompassing several approaches to telescope design, focal plane design, scanning system design, data processing system design, and orbits/coverage and operations. The resulting sensor system designs are highly capable, compared to existing systems, meet the performance goals established, and yet fit within the tactical satellite class.