รายละเอียดข้อมูลการตีพิมพ์เผยแพร่
ปีพ.ศ./ กศษ./ ปงม.
2566 / 2565 / 2566
สาขาการตีพิมพ์/เผยแพร่
มนุษยศาสตร์และสังคมศาสตร์
ประเภทการตีพิมพ์/เผยแพร่
วารสารวิชาการระดับนานาชาติ Scopus Q2
ชื่อเรื่อง/ผลงานการตีพิมพ์ฯ
A hybrid HS-LS conjugate gradient algorithm for unconstrained optimization with applications in motion control and image recovery
จากโครงการวิจัยเรื่อง
ผู้นำเสนอผลงาน
แหล่งตีพิมพ์ฯ
Journal of Computationa and Applied Mathematics, 433(2023), 115304.
ISSN: 0377-0427
(ฐานข้อมูลนานาชาติ Scopus, SJR Q2)
บทคัดย่อ
This article presents a new hybrid conjugate gradient (CG) algorithm for solving un-constrained optimization problem. The search direction is defined as a combination of
Hestenes–Stiefel (HS) and the Liu–Storey (LS) CG parameters and is close to the direction
of the memoryless Broyden–Fletcher–Goldferb–Shanno (BFGS) quasi-Newton direction.
In addition, the search direction is descent and bounded. The global convergence of the
algorithm is obtained under the Wolfe-type and Armijo-type line searches. Numerical
experiments on some benchmark test problems is carried out to depict the efficiency and
robustness of the hybrid algorithm. Furthermore, a practical application of the algorithm
in motion control of robot manipulator and image restoration is provided.
คำสำคัญ
Unconstrained optimization,Three-term conjugate gradient method,Line search,Global convergence
วันที่นำเสนอ
1 มกราคม 2566   -   31 ธันวาคม 2566
สัดส่วนคะแนน
1.00
เอกสารดาวน์โหลด
ดาวน์โหลด   ดาวน์โหลด 16 ครั้ง
หมายเหตุ
--