An Energy-Aware Multilayer Clustering-Based Butterfly Optimization Routing for Underwater Wireless Sensor Networks

Wireless Personal Communications - Underwater wireless sensor network transmission leads to higher energy consumption due to propagation distance and high energy loss. Moreover, the acoustic medium...     

An energy-efficient distributed node clustering routing protocol with mobility pattern support for underwater wireless sensor networks

Wireless Networks - Underwater Wireless Sensor Networks (UWSN) has received more attention in exploring promising technologies for scientific data collection of underwater natural resources with...     

Electrochemical machining process parameter optimization using particle swarm optimization

Electrochemical machining (ECM) is a nontraditional process used for the machining of hard materials and metal‐matrix composites. Composites are used in several applications such as aerospace, automobile industries, and medical field. The determination of optimal process parameters is difficult in the ECM process for obtaining maximum material removal rate (MRR) and good surface roughness (SR). In this paper, a multiple regression model is used to obtain the relationship between process parameters and output parameters. Particle swarm optimization (PSO) is one of the optimization techniques for solving the multiobjective problem; it is proposed to optimize the ECM process parameters. Current (C), voltage (V), electrolyte concentration (E), and feed rate (F) are considered as process parameters, and MRR and SR are the output parameters used in the proposed work. The developed multiple regression is statistically analyzed by regression plot and analysis of variance. The optimized value of MRR and SR obtained in PSO is 0.0116 g/min and 2.0106 μm, respectively. Furthermore, PSO is compared with the genetic algorithm. The PSO technique outperforms the genetic algorithm in computation time and statistical analysis. The obtained values are validated to test the significance of the model, and it is noticed that the error value for MRR and SR is within 0.15%.