Design and optimisation of microelectromechanical systems: a review of the state-of-the-art

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

“…One of the main challenges in the development of MEMS devices is their design and optimisation [9]. Traditional development of MEMS by silicon micromachining fabrication techniques [10] requires multiple prototypes and extensive experimentation (design process).…”

“…This process is often based around the parameterization of a MEMS device, which is modelled and then followed by a hand-driven optimisation routine. This allows the designer to input various sets of design variables in order to optimise the device, leading to eventual fabrication, testing and ultimately re-evaluation and redesign [9]. This trial and error approach is dependent on the designers experience and an antithesis to the goal of allowing designers the ability to focus on device and system performance.…”

“…The development of a design optimisation environment, which can allow MEMS designers to automate the process of modelling, simulation and optimisation at all levels of the MEMS design process, is beneficial to the progress of the MEMS Industry. Such an environment reduces the burden on the designer and provides a background that will potentially produce optimal devices within the design constraints [9,13,14]. Though traditional methods dominate in the automated design optimisation of MEMS, there are current areas of optimisation which show promise in overcoming some of the limitations within local gradient-based search by incorporating a more stochastic approach to optimisation.…”

“…Other areas exploiting MOGAs for MEMS design optimisation include hierarchical or multi-level design [14,27], robust design methodologies [60] and hybrid genetic algorithms [62]. Different researchers have used different classes of EAs in this subject domain [9,26,27], however it is not clear which particular EA approach is the most appropriate and efficient in MEMS design synthesis and optimisation. This paper presents a novel framework for the multi-objective design optimisation of MEMS which is then used as a foundation for a comparative investigation into the performance of two widely used EAs (Multi-Objective Genetic Algorithm: MOGA-II [28] and NSGA-II [29]) through different MEMS design optimisation case studies.…”

Automatic synthesis of MEMS devices using self-adaptive hybrid metaheuristics

“…One reason for this is the complexity involved in the modeling, design and fabrication of MEMS. There are many constraints in designing and fabricating MEMS devices due to the limitations of current fabrication techniques [1,3]. However, as process technologies become more stable, research emphasis can be shifted from developing specific process technologies towards the design of systems with a large number of reusable components, such as resonators, accelerometers, gyroscopes, and micro-mirrors.…”