Sizing optimization of truss structures using the political optimizer (PO) algorithm

The challenge that structural optimization tasks pose by dint of their highly non-linear design constraints and complex solution domains has spurred both the implementation and development of many novel metaheuristic algorithms over the past two decades. In this study, the very recently proposed political optimizer algorithm has been implemented in the context of structural optimization. Inspired by the multi-phased political process in parliamentary democracies, the algorithm efficiently balances between exploration and exploitation by logically dividing the population of search agents into political parties that compete for constituency (electoral district) domination. This competitive-based population partitioning scheme ensures that enough of the search space is thoroughly investigated for the global optima all the while maintaining algorithmic speed and efficiency. Moreover, a local search mechanism known as the recent past-based position updating strategy (RPPUS) provides for an effective exploitation strategy where every search agent is allowed to learn from its previous behavior and hence collectively guide the population towards better solutions and away from worse ones. To quantitatively assess the performance of the algorithm, three planar trusses (10 bar, 18 bar, and 200 bar) and four space trusses (22 bar, 25 bar, 72 bar, and 942-bar) with multiple loading conditions and design constraints have been considered. Results show that for small/medium-scale structural systems, the PO algorithm outperforms all previously proposed state-of-the-art optimization methodologies in all aspects may it be final optimized weight, algorithmic stability, or convergence speeds, and that for larger structures excellent performance is still maintained but a certain, yet acceptable, extent of algorithmic instability is manifest. Based off these findings, future research into the PO algorithm as an efficient structural optimizer is strongly recommended.