Stochastic search-based actuator placement optimization for adaptive beam string structures

doi:10.1016/j.istruc.2024.108087

2025-01-01

PRE

DOI：10.1016/j.istruc.2024.108087

原文链接

原文求助

摘要

En 中文

By replacing the vertical struts in a traditional beam string structure with actively retractable actuator struts, the deformation and stress of the structure can be proactively adjusted in response to external environmental conditions, thereby forming the adaptive beam string structure (ABSS). This study proposes an optimization method for the actuator strut placement in ABSS, with the objective of enhancing the control efficiency of the actuating system while simultaneously reducing the construction and maintenance costs. Specifically, a constrained simulated annealing algorithm (CSA) is used to obtain the optimal placement scheme, integrated with a heuristic strategy based on sensitivity analysis of the actuator struts to improve search efficiency. A novel hybrid genetic algorithm (HGA) is proposed to efficiently address the adaptive control optimization problem of ABSS, enabling evaluation of various schemes and providing feedback to the CSA. The numerical simulations across three cases validate the effectiveness of the proposed optimization method. With approximately one-third of the actuators removed, Case 1 retains 69 % of the original control performance, while the more complex Case 2 retains 94 %. In the most complex Case 3, the control performance retains 98 % with half of the actuators removed. These results demonstrate that the proposed method effectively reduces the number of actuators in ABSS while maintaining the adaptive control performance without significant decline, thereby enhancing overall control efficiency. The improvement becomes more significant with increased structural complexity, indicating the potential of actuator placement optimization in complex ABSS.

Keyword:

Beam string structure

Adaptive structure

Actuator optimization

Hybrid genetic algorithm

Constrained simulated annealing algorithm