Optimum design of high-dimensional truss structures using the bonobo optimizer with inertia weight strategies

Abstract

This study investigates the effectiveness of the Improved Bonobo Optimizer with Inertia Weight (IWBO) in optimizing various large-scale truss structures, including complex 345, 336 and 284-bar trusses with sizing considerations, as well as a 260-bar truss, incorporating both sizing and layout variables. To enhance the global and local exploration capabilities of the Bonobo Optimizer (BO), nine different inertia weight strategies-Butterworth, simulated annealing, exponential decreasing, logarithmic decreasing, sigmoid decreasing/increasing, oscillating, linearly decreasing and nonlinear time-variant inertia weights-were implemented and evaluated to identify the optimal strategy. To evaluate the mathematical performance of the IWBO, the CEC2021 benchmarks were used along with 10 functions characterized by biased, rotated and shifted properties. The experimental results are highly promising, indicating that the proposed approach demonstrates robust performance under challenging optimization conditions and may serve as a compelling alternative for complex optimization problems. The OSC-based BO consistently outperforms competing methods, demonstrating superior convergence to the lowest fitness value across all test cases. Experimental results reveal that the IWBO enhances the standard BO, achieving lower values for key statistical parameters. This study underscores the effectiveness of the IWBO in optimizing large-scale truss structures, highlighting its potential to offer improvements in efficiency for structural engineering applications.