Search

A methodology is presented for the design and implementation of robust controllers that optimise inter-sample performance for a broad range of nonlinear systems operated within a sampled-data environment. The methodology applies to nonlinear continuous-time systems described by state equations, and it allows for modelling uncertainties and constraints on maximal control effort. It is shown that there exist optimal robust state-feedback controllers that minimise inter-sample tracking errors for such systems, as long as an appropriate controllability requirement is met. A relatively simple design and implementation procedure for such controllers is described. [ABSTRACT FROM AUTHOR]

The problem of controlling a finite-state asynchronous sequential machine is examined. Main consideration is given to input/output control, where access to the state of the machine is not available. The objective is to use output feedback to control the machine so as to match a prescribed model. It is shown that necessary and sufficient conditions for the existence of appropriate controllers can be stated in terms of a simple F comparison of two numerical matrices. Whenever controllers exist, algorithms for their design are outlined. Index Terms--Asynchronous sequential machines, model matching, observers, output feedback, separation principle.

The problem of eliminating the effects of critical races on asynchronous machines is considered in a control theoretic context. State feedback controllers that eliminate the effects of critical races are developed. The results include necessary and sufficient conditions for the existence of such controllers and algorithms for their design. When the controllers exist, they eliminate the race effects and control the machine to match a given race-free model. Index Terms--Asynchronous machines, feedback, finite state machines, hazards, races.

The existence and implementation of optimal robust state-feedback controllers that achieve optimal tracking over the infinite horizon is considered. The objective is to minimise over all times the deviation of a nonlinear system from a specified target state. It is shown that optimal robust state-feedback controllers that achieve this objective exist for a broad class of nonlinear input-affine systems. It is also shown that optimal performance can be approximated by state-feedback controllers that are relatively easy to design and implement. [ABSTRACT FROM AUTHOR]

This paper deals with the control of a nonlinear system whose output is subjected to an additive disturbance. The main result is a simple parameterization of the set of all system responses that can be obtained through internally stable control of the given system. The parameterization provides a clear indication of the effects of the disturbance on the response of the stabilized closed loop system. The class of achievable responses is determined by the 'numerator' of a right coprime fraction representation of the system being controlled.

The use of Talbot filters in improving the performance of laser diode arrays is studied. Laboratory tests are conducted on an array of 10 equally-spaced emitters to evaluate the efficiency of opaque Talbot filters in controlling the lasers' in-phase and out-of-phase lateral mode operations under continous wave conditions. Results indicate that spatial Tabot filtering can significantly improve mode control in high-power laser diode arrays.

The design of optimal robust state-feedback controllers that guide a system to a target in minimal time is considered under constraints on the maximal input amplitude and the maximal overshoot of the controlled system. It is shown that such robust feedback controllers exist for a rather broad family of nonlinear systems. It is also shown that optimal performance can be approximated by state-feedback controllers that are relatively easy to design and implement. [ABSTRACT FROM AUTHOR]

The process of periodic sampling is investigated for a class of nonlinear systems. The objective is to achieve the longest sampling period that is compatible with a specified error bound. It is shown that there are robust optimal controllers that achieve this objective. It is also shown that the performance of such optimal controllers can be approximated by bang-bang controllers – controllers that are relatively easy to design and implement. [ABSTRACT FROM AUTHOR]

Resonant-layer effect (RLE) optical isolators can be integrated with a wide variety of optical waveguide devices and systems and theoretically provide a high degree of isolation with low insertion loss. The effect of mode coupling within the isolator and the losses due to coupling between the isolator and the input and output waveguides on the performance of RLE isolators is calculated. The results predict that isolators integrated with semiconductor waveguides similar to those used for efficient lasers and amplifiers and isolators integrated with quartzlike waveguides can give 40 dB of isolation with insertion losses below 3 dB. The calculations also show relaxed dimensional tolerance and lengths in the 1–2 mm range, which makes commercial production of such isolators attractive. [ABSTRACT FROM AUTHOR]