Computing upper bounds of the positive real roots of some polynomials is a key step of those real root isolation algorithms based on continued fraction expansion and Vincent's theorem.The authors give a new algori...Computing upper bounds of the positive real roots of some polynomials is a key step of those real root isolation algorithms based on continued fraction expansion and Vincent's theorem.The authors give a new algorithm for computing an upper bound of positive roots in this paper.The complexity of the algorithm is O(n log(uH-l))additions and multiplications where u is the optimal upper bound satisfying Theorem 3.1 of this paper and n is the degree of the polynomial.The method together w辻h some tricks have been implemented as a software package logcf using C language.Experiments on many benchmarks show that logcf is competitive with Root Intervals of Mathematica and the function realroot of Maple averagely and it is much faster than existing open source real root solvers in many test cases.展开更多
The paper is concerned with the improvement of the rational representation theory for solving positive-dimensional polynomial systems. The authors simplify the expression of rational representation set proposed by Tan...The paper is concerned with the improvement of the rational representation theory for solving positive-dimensional polynomial systems. The authors simplify the expression of rational representation set proposed by Tan and Zhang(2010), obtain the simplified rational representation with less rational representation sets, and hence reduce the complexity for representing the variety of a positive-dimensional ideal. As an application, the authors compute a "nearly" parametric solution for the SHEPWM problem with a fixed number of switching angles.展开更多
为了解决当前存在的区间运算复杂性高、普通区间迭代程序运算量大、花费机器时间长等问题,提高区间迭代的运算效率,针对单变元多项式方程的求解展示了一个高效的区间Newton迭代算法。该算法利用1维问题多维化的思想加快迭代的收敛速度,...为了解决当前存在的区间运算复杂性高、普通区间迭代程序运算量大、花费机器时间长等问题,提高区间迭代的运算效率,针对单变元多项式方程的求解展示了一个高效的区间Newton迭代算法。该算法利用1维问题多维化的思想加快迭代的收敛速度,改进了多维化过程中出现的性能拖累,极大地提高了算法在实际应用中的执行效率。算法已经被实现为M ap le程序,实验数据表明,与现有的算法相比,这个算法的迭代次数和运行时间都大幅减少,充分显示了它的高效性。展开更多
基金supported by the National Science Foundation of China under Grant Nos.61802318,61732001and 61532019
文摘Computing upper bounds of the positive real roots of some polynomials is a key step of those real root isolation algorithms based on continued fraction expansion and Vincent's theorem.The authors give a new algorithm for computing an upper bound of positive roots in this paper.The complexity of the algorithm is O(n log(uH-l))additions and multiplications where u is the optimal upper bound satisfying Theorem 3.1 of this paper and n is the degree of the polynomial.The method together w辻h some tricks have been implemented as a software package logcf using C language.Experiments on many benchmarks show that logcf is competitive with Root Intervals of Mathematica and the function realroot of Maple averagely and it is much faster than existing open source real root solvers in many test cases.
基金supported by the National Natural Science Foundation of China under Grant No.11671169Scientific Research Fund of Liaoning Provincial Education Department under Grant No.L2014008
文摘The paper is concerned with the improvement of the rational representation theory for solving positive-dimensional polynomial systems. The authors simplify the expression of rational representation set proposed by Tan and Zhang(2010), obtain the simplified rational representation with less rational representation sets, and hence reduce the complexity for representing the variety of a positive-dimensional ideal. As an application, the authors compute a "nearly" parametric solution for the SHEPWM problem with a fixed number of switching angles.
文摘为了解决当前存在的区间运算复杂性高、普通区间迭代程序运算量大、花费机器时间长等问题,提高区间迭代的运算效率,针对单变元多项式方程的求解展示了一个高效的区间Newton迭代算法。该算法利用1维问题多维化的思想加快迭代的收敛速度,改进了多维化过程中出现的性能拖累,极大地提高了算法在实际应用中的执行效率。算法已经被实现为M ap le程序,实验数据表明,与现有的算法相比,这个算法的迭代次数和运行时间都大幅减少,充分显示了它的高效性。
