Root Based Optimization Algorithm for Task-Oriented Structural Design of a Multi-Axial Road Test Rig

Abstract

A novel root based optimization algorithm (ROA), which mimics the proliferation of plant roots, is proposed to evaluate the performances of a multi-axial road test rig over a prescribed workspace. Plant roots have developed complex behavior patterns to search for nutrients in the soil and even exhibit swarm intelligence. The growing roots can adapt their actions such as elongation, branching, and tropistic movement according to the environment. During the process of foraging for water and nutrients, the adaptive growth behavior of roots can be simulated to tackle the optimization problem. The efficiency of the proposed ROA has been validated by comparing it with well-known intelligent algorithms using classical and IEEE CEC 2019 benchmark functions. Computational results indicate that ROA outperforms other algorithms to search a global optimum on several benchmarks. Furthermore, the results are tested with nonparametric statistical methods, e.g., Wilcoxon rank-sum test. Besides, ROA is applied to the optimal dimensioning synthesis of the multi-axial road test rig (MRTR) based on the orthogonal 6-RSS parallel manipulator to promote the device's performance. The inverse kinematics of the multi-axial road test rig based on the asymmetrical orthogonal 6-RSS parallel mechanism is analyzed in detail. Applying the achieved dimensionless Jacobian, the performance criteria involving the global isotropy index, the force/velocity transmissibility index, the decoupling index are established. By simultaneously taking account the proposed criteria of performance, structural optimization is carried out resorting to ROA. Finally, a set of optimized dimensional parameters is obtained. The practical application further illustrates that ROA has the potential to deal with complex optimization problems.

Hybrid Sine Cosine and Fitness Dependent Optimizer for Global Optimization

“…Nature-inspired algorithms include particle swarm optimization (PSO) [1], differential evaluation (DE) [2] and genetic algorithm (GA) [3]. Some of the recent nature-inspired algorithms are sine cosine algorithm (SCA) [4] [5], fitness dependent optimizer (FDO) [6], wingsuit flying search (WFS) [7], whale optimization algorithm (WOA) [8] [9], butterfly optimization algorithm (BOA) [10] [11], dragonfly algorithm (DA) [12] [13], grey wolf optimizer (GWO) [14], moth-flame optimization algorithm (MFO) [15]- [18], root-based VOLUME XX, 2017 1 optimization algorithm [19], coot algorithm [20] and colony predation algorithm [21]. This paper focuses on the fitness-dependent optimizer (FDO) proposed by Abdullah and Rashid [6].…”