Program Package for Process Identification and Control System Design

Popescu, Theodor D.; Sima, Vasile; Varga, A.; Vasiliu, C.

doi:10.1016/b978-0-08-024488-4.50097-8

Cited by 1 publication

(1 citation statement)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The SLICOT library is the result of international cooperation among leading numerical algebra and control numer- [23]. BIMAS [24] extends the capabilities of the underlying EISPACK and LINPACK software for controlrelated computational problems, such as solving matrix equations (Lyapunov, Sylvester, Riccati), computing matrix exponentials, and reordering Schur forms.…”

Section: Control Librariesmentioning

confidence: 99%

High-performance numerical software for control

et al. 2004

View full text Add to dashboard Cite

his article is concerned with the development of quality control-application software that performs efficiently and reliably on modern computing machines. In particular, we describe the subroutine library SLICOT (Subroutine Library In COntrol Theory) [1], give the background and motivation for the development of this library, and describe its most important features. The need to solve the increasingly challenging large-scale problems that arise in control systems analysis and design has led to a need for more powerful algorithms able to solve these problems. Most users of standard control-oriented commercial software tools tend to have a blind confidence in the "quality" of the computed solutions and in the "efficiency" of the underlying methods. In many cases, however, it can be difficult to judge the numerical accuracy of the solutions as well as the efficiency of the implemented methods. It is a sad and mostly ignored fact that in many cases such tools are of dubious numerical quality. Therefore, to judge the numerical performance of algorithms, Developing reliable and efficient software for control applications. February 2004 62 IEEE Control Systems Magazine acceptably well-conditioned problems. On the other hand, unstable algorithms could produce wrong results even for well-conditioned problems.

show abstract

Section: Control Librariesmentioning

confidence: 99%