Multi-objective spur gear design using teaching learning-based optimization and decision-making techniques

Multi-objective optimization of a 2-stage spur gearbox using NSGA-II and decision-making methods

“…As shown in Table 3, the ideal values for power output, volume and centre-distance are 9.956 kW, 894,036.77 mm 3 and 191.25 mm while nadir values are 9.933 kW, 2,479,462.66 mm 3 and 329.26 mm 3 . Considering volume and power output only, as shown in Table 4, the ideal values were 893,006.7 mm 3 and 9.97 kW while nadir values were 2,470,024.28 mm 3 and 9.93 kW.…”

“…To extract optimum results from the frontier, decision methods are applied such as discussed in the literature [3]. In this section, the NSGA-II algorithm is discussed in Sect.…”

“…However, before application, it is required to unify the scale and dimensions of all the objective functions consistently. To achieve this, objectives must be dimensionless and this can be achieved through the Euclidean, linear and fuzzy methods [3,4]. Concerning these methods, F ij represents the objectives matrix reflecting the Pareto solutions and i is the index of points on the Pareto frontiers and j represents each objective in the optimization space.…”

“…In this method, the solitary keys are chosen from the Pareto frontier. The intersection of constraint sets and fuzzy variables results in membership functions [3]. Therefore, at every unique upshot, the individual can go for a fuzzy route for the minimum figure of the membership function.…”

“…The advancement of technology has led to increased accessibility to computing resources and hence the introduction of meta-heuristic methods viz. Genetic algorithm (GA), simulated annealing (SA), ant-colony optimization (ACO) and teaching and learning-based optimization (TLBO) [3,4] came into the picture. Compared to heuristic methods, which are problem-dependent and efficient in local search, meta-heuristic methods achieve global optimal solutions by applying a superior search strategy which is informed by feedback from previous solutions, objective and algorithm performance [5,6].…”