A criterion for root exclusion from a region composed as a union of elemental regions-discs and halfplanes-is established. Associated with each elemental region is a derived polynomial which must be by the criterion a strictly Hurwitz polynomial. The criterion is the basis for a root exclusion test,
Polynomial Root Clustering
β Scribed by T.A. Bickart; E.I. Jury
- Publisher
- Elsevier Science
- Year
- 1979
- Tongue
- English
- Weight
- 536 KB
- Volume
- 308
- Category
- Article
- ISSN
- 0016-0032
No coin nor oath required. For personal study only.
β¦ Synopsis
A sequence of tests on derived polynomials to be strictly Hurwitz polynomials is shown to be equivalent to a given (typically real) polynomial having all its zeros
in an open sector, symmetric with respect to the real axis, in the left half-plane. 7'he number of tests needed is at most 1 + [(In k)/(ln 3)1, w h ere k is the integer associated with the central angle r/k of the sector. An extension of this result on the sector as a region of root clustering is given which shows that only a limited number of tests are needed to verify that the roots are clustered in a region composed as the intersection of a set of primative (sector-like) regions. The results reported evolve from application of a collection of mappings on the complex plane defined by a particular collection of Schwarz-Christoflel transformations.
π SIMILAR VOLUMES
There is a well-known lower bound, due to Mignotte, for the minimum root separation of a squarefree integral polynomial, but no evidence for the sharpness of this bound. This paper provides massive computational evidence for a conjectured much larger bound, one that is approximately the square root
## We restrict the discussion of root clustering (exclusion) to rational convex regions. A region in thisfamily is constructed as the intersection of an injinite number of halfplanes. The root clustering criterion inooloes the positiuity ofa set ofpolynomials with respect to the region's parameter
A new procedure is proposed for interpolation and extrapolation of functions by a root of a low=degree polynomial. Three examples of the application of the procedure are presented. The rms error in representing SCF surfaces ranges from 2 cal/mole (linear HCN) to 0.23 kcal/mole (C2v MgH2 triplet).