1. A Novel Method of Dynamic Force Identification and Its Application
- Author
-
Qinhui Liu, Nengjian Wang, Chunping Ren, and Chunsheng Liu
- Subjects
021103 operations research ,Article Subject ,Computer science ,lcsh:Mathematics ,General Mathematics ,Work (physics) ,0211 other engineering and technologies ,General Engineering ,Structure (category theory) ,Stability (learning theory) ,010103 numerical & computational mathematics ,02 engineering and technology ,Construct (python library) ,lcsh:QA1-939 ,01 natural sciences ,Identification (information) ,lcsh:TA1-2040 ,Conjugate gradient method ,Force dynamics ,0101 mathematics ,lcsh:Engineering (General). Civil engineering (General) ,Algorithm - Abstract
In this paper, an efficient mixed spectral conjugate gradient (EMSCG, for short) method is presented for solving unconstrained optimization problems. In this work, we construct a novel formula performed by using a conjugate gradient parameter which takes into account the advantages of Fletcher–Reeves (FR), Polak–Ribiere–Polyak (PRP), and a variant Polak-Ribiere-Polyak (VPRP), prove its stability and convergence, and apply it to the dynamic force identification of practical engineering structure. The analysis results show that the present method has higher efficiency, stronger robust convergence quality, and fewer iterations. In addition, the proposed method can provide more efficient and numerically stable approximation of the actual force, compared with the FR method, PRP method, and VPRP method. Therefore, we can make a clear conclusion that the proposed method in this paper can provide an effective optimization solution. Meanwhile, there is reason to believe that the proposed method can offer a reference for future research.
- Published
- 2019