รหัสดีโอไอ 10.14457/TU.the.2019.744
Title Truss topology optimization with a prescribed maximum number of elements
Creator Alin Shakya
Contributor Pruettha Nanakorn, Advisor
Publisher Thammasat University
Publication Year 2019
Keyword Truss topology optimization ,Coding scheme ,Particle swarm optimization ,Unwanted elements ,Ground structure ,Multi-population.
Abstract Truss topology optimization is one of the most active research areas in the field of structural optimization. Many studies on truss topology optimization have been done. Most of these studies employ the ground structure approach in performing truss topology optimization. In the ground structure approach, a grid of nodes is created, and a dense ground structure is usually formed by joining every node of the grid. The elements in the ground structure are then used as possible positions of truss elements. Since the elements are created by joining all the nodes of the grid, the search space becomes large quickly when the number of nodes is increased. In truss topology optimization problems, there are commonly many kinematically unstable trusses in their search spaces. Oftentimes, trusses are unstable only because of some locally unstable truss elements, such as disconnected elements. In addition, many trusses in a search space contain useless zero-force elements. Many of these kinematically unstable and zero-force elements can be easily identified. In the first part of this study, a new representation for truss topology optimization is proposed in which an element-removal algorithm is used to remove such types of unwanted elements from trusses. These elements are removed during the translation of a representation code into its corresponding truss in order that the resulting truss may become a stable one. As a result, more representation codes in the search space are mapped into kinematically stable and efficient trusses, and the level of competition among representation codes is increased. This process alleviates the problem of having large search spaces and encourages faster solution convergences. In the second part of the study, a new coding scheme is proposed in which a prescribed maximum number of elements is specified for truss optimization. In this scheme, there is no need for any ground structure, and only a grid of nodes is required. The value of the maximum number of elements is set to be lower than the number of elements in a complete ground structure. As a result, the number of elements that must be removed during the optimization process to obtain the optimal truss is reduced. Another advantage of the proposed scheme is that the designer can predefine a maximum limit for the number of members to be present in the final optimal truss. Both proposed representations are used in this study with a simple multi-population particle swarm optimization algorithm (MPSO). Several example problems are solved, and it is found that both proposed representation codes significantly improve the performance of the optimization process.
ดิจิตอลไฟล์ Digital File #1

บรรณานุกรม

Alin Shakya และผู้แต่งคนอื่นๆ. (2019) Truss topology optimization with a prescribed maximum number of elements. Thammasat University:ม.ป.ท.
Alin Shakya และผู้แต่งคนอื่นๆ. 2019. Truss topology optimization with a prescribed maximum number of elements. ม.ป.ท.:Thammasat University;
Alin Shakya และผู้แต่งคนอื่นๆ. Truss topology optimization with a prescribed maximum number of elements. ม.ป.ท.:Thammasat University, 2019. Print.