An efficient evolutionary multi-objective framework for MEMS design optimisation: validation, comparison and analysis

Abstract

The application of multi objective evolutionary algorithms (MOEA) in the design optimisation of microelectromechanical systems (MEMS) is of particular interest in this research. MOEA is a class of soft computing techniques of biologically inspired stochastic algorithms, which have proved to outperform their conventional counterparts in many design optimisation tasks. MEMS designers can utilise a variety of multi-disciplinary design tools that explore a complex design search space, however, still follow the traditional trial and error approaches. The paper proposes a novel framework, which couples both modelling and analysis tools to the most referenced MOEAs (NSGA-II and MOGA-II). The framework is validated and evaluated through a number of case studies of increasing complexity. The research presented in this paper unprecedentedly attempts to compare the performances of the mentioned algorithms in the application domain. The comparative study shows significant insights into the behaviour of both of the algorithms in the design optimisation of MEMS. The paper provides extended discussions and analysis of the results showing, overall, that MOGA-II outperforms NSGA-II, for the selected case studies.

Multi-level and multi-objective design optimisation of a MEMS bandpass filter

“…Early work done by Zhou [ 11 ] looked to apply multi-objective genetic algorithms to the design and optimisation of a number of device level MEMS devices using a NODAL simulator [ 12 ] and over the years the use of evolutionary computational methods in MEMS design has expanded into a number of areas such as conceptual design [ 13 ], component based design, classical shape [ 14 ], sizing [ 15 ] and topological [ 16 ] design optimisation to interactive [ 17 ] and casebased reasoning methods [ 18 ]. A comprehensive list of conventional and unconventional methods for MEMS design optimisation can be found in [ 19 ]. Hierarchical or multi-level methods are another unconventional approach which looks to utilize their specific architecture to benefit the design optimisation process.…”

A Fast Multiobjective Optimization Strategy for Single-Axis Electromagnetic MOEMS Micromirrors

“…Consequently, structured, Computer-Aided-Design (CAD) assisted approaches to MEMS/MOEMS design optimization would be welcome. There is, however, evidence that MEMS design is still largely based on inherently suboptimal ad-hoc, trial-and-error methods [ 16 , 17 , 18 , 19 ]. Part of this problem can be attributed to the lower maturity of the MEMS field with respect to other fields of engineering, which also reflects in a lack of efficient integrated tools for optimal design.…”

“…They used an analytical model for the output noise and determined the Pareto frontier with respect to two objectives (the noise spectral density and the area occupancy of the device) with a multiobjective evolutionary algorithm based on decomposition (MOEA/D) [ 23 ]. Farnsworth et al [ 19 ] developed a flexible system that applies MO through evolutionary algorithms. The system interfaces with CAD packages or user-provided scripts for device description, and different case studies for either option are presented.…”

Multi co-objective evolutionary optimization: Cross surrogate augmentation for computationally expensive problems

“…Since the design cycle time of a product is directly proportional to the number of calls made to the costly analysis solvers, researchers are seeking for novel multi-objective optimization frameworks that can handle these forms of problems elegantly. Besides parallelism, which is an obvious choice to achieving near linear order improvement in evolutionary search, researchers are gearing towards surrogateassisted or meta-model assisted evolutionary frameworks when handling optimization problems imbued with costly non-linear objective and constraint functions [13], [14], [15], [16], [17].…”