Penalty Function Based Critical Point Approach to Compute Real Witness Solution Points of Polynomial Systems
We present a critical point method based on a penalty function for finding certain solution (witness) points on real solutions components of general real polynomial systems. Unlike other existing numerical methods, the new method does not require the input polynomial system to have pure dimension or satisfy certain regularity conditions.
This method has two stages. In the first stage it finds approximate solution points of the input system such that there is at least one real point on each connected solution component. In the second stage it refines the points by a homotopy continuation or traditional Newton iteration. The singularities of the original system are removed by embedding it in a higher dimensional space.
In this paper we also analyze the convergence rate and give an error analysis of the method. Experimental results are also given and shown to be in close agreement with the theory.
The authors would like to thank the anonymous reviewers for their constructive comments that greatly helped improving the paper. This work is partially supported by the projects NSFC (11471307, 11671377, 61572024), cstc2015jcyjys40001, and the Key Research Program of Frontier Sciences of CAS (QYZDB-SSW-SYS026).
- 17.Hong, H.: Improvement in CAD-Based Quantifier Elimination. Ph.D. thesis. Ohio State University, Columbus, Ohio (1990)Google Scholar
- 18.Li, T.Y., Lee, T.L.: Homotopy method for solving Polynomial Systems software. http://www.math.msu.edu/~li/Software.htm
- 22.Lu, Y.: Finding all real solutions of polynomial systems. Ph.D thesis. University of Notre Dame (2006). Results of this thesis appear. In: (with Bates, D.J., Sommese, A.J., Wampler, C.W.), Finding all real points of a complex curve, Contemp. Math. vol. 448, pp. 183–205 (2006)Google Scholar
- 23.Ma, Y., Zhi, L.: Computing Real Solutions of Polynomial Systems via Low-rank Moment Matrix Completion. In: ISSAC, pp. 249–256 (2012)Google Scholar
- 25.El Din, M.S., Schost, É.: Polar varieties and computation of one point in each connected component of a smooth real algebraic set. In: ISSAC 2013, pp. 224–231 (2003)Google Scholar
- 27.Sommese, A.J., Wampler, C.W.: The Numerical Solution of Systems of Polynomials Arising in Engineering and Science. World Scientific Press (2005)Google Scholar
- 28.Stewart, G.W.: Perturbation theory for the singular value decomposition. In: SVD and Signal processing, II: Algorithms, Analysis and Applications, pp. 99–109. Elsevier (1990)Google Scholar
- 31.Wu, W., Reid, G.: Finding points on real solution components and applications to differential polynomial systems. In: ISSAC, pp. 339–346 (2013)Google Scholar
- 32.Wu, W., Reid, G., Feng, Y.: Computing real witness points of positive dimensional polynomial systems. Accepted by Theoretical Computer Sciences (2017). http://doi.org/10.1016/j.tcs.2017.03.035
- 33.Yang, Z., Zhi, L., Zhu, Y.: Verified error bounds for real solutions of positive-dimensional polynomial systems. In: ISSAC, pp. 371–378 (2013)Google Scholar