Conditions for removing intersample ripples in multirate control

Abstract

Multirate systems arise when signals of interest are sampled at d i f f m n t rates. Measurements jivm chem-ical processes are typically available at different sampling rates. For example, composition estimates in a distillation column are available at a much slower rate than pow, temperature and pressure measurements. Multirate s y s t e m pose a challenging problem due to several reasons such as inmmed complexity design, time-varying nature, etc.. Systems consisting of fast-rate control moves and slowsampled outputs are a common scenario in chemical processes and of practical interest. lhditionally, inferential techniques based o n secondary measurements have been used to design the fast-rate input moves. Lifting techniques conveniently transform multirate systems to single-rate lifted systems with i n c m e d dimensionality. Controllers designed using lifting techniques require that certain cawality constmints are satisfied. I n this woTk, we show firstly, that intersample ripples can arise in the closed-loop output of a multirate system as a result of non-uniform gains of the discrete lifted system and inverse lifting. Secondly, we present conditions on the controller gains to avoid intersample ripples.

Optimal flight control design in irregular sampled-data control systems

“…The output variable should be considered slowly measured. Basically, there are two different methodologies for solving this problem: inferential [1], [2] and sampled-data control [3], [4] techniques. The sampled-data control systems are a special case of periodic systems that the sampling scheme repeats over a global period which is equal to the least common multiple of sampling periods of all system signals or time-varying systems, and due it many concepts and ideas of them are applicable to irregular sampled-data systems [5][6][7][8].…”