Optimized Design of a Self-Biased Amplifier for Seizure Detection Supplied by Piezoelectric Nanogenerator: Metaheuristic Algorithms versus ANN-Assisted Goal Attainment Method

Abstract

This work is dedicated to parameter optimization for a self-biased amplifier to be used in preamplifiers for the diagnosis of seizures in neuro-diseases such as epilepsy. For the sake of maximum compactness, which is obligatory for all implantable devices, power is to be supplied by a piezoelectric nanogenerator (PENG). Several meta-heuristic optimization algorithms and an ANN (artificial neural network)-assisted goal attainment method were applied to the circuit, aiming to provide us with the set of optimal design parameters which ensure the minimal overall area of the preamplifier. These parameters are the slew rate, load capacitor, gain–bandwidth product, maximal input voltage, minimal input voltage, input voltage, reference voltage, and dissipation power. The results are re-evaluated and compared in the Cadence 180 nm SCL environment. It has been observed that, among the metaheuristic algorithms, the whale optimization technique reached the best values at low computational cost, decreased complexity, and the highest convergence speed. However, all metaheuristic algorithms were outperformed by the ANN-assisted goal attainment method, which produced a roughly 50% smaller overall area of the preamplifier. All the techniques described here are applicable to the design and optimization of wearable or implantable circuits.

A Comprehensive Review of Bio-Inspired Optimization Algorithms Including Applications in Microelectronics and Nanophotonics

“…Our own experience is that some approaches may prove excellent in one type of problem and very poor in another, sometimes unexpectedly so. One such example out of many was our optimization of an analog preamplifier circuit area for early epileptic seizure detection which was designed better by applying an ANN-assisted goal attainment method than by attempting to independently apply several well-known metaheuristic algorithms [ 231 ]. Useful tools regarding this problem are the existing sets of benchmark tests which help uncover some advantages and disadvantages of algorithms, as well as the publications that give results of their applications [ 81 , 82 , 83 , 97 ].…”

“…This helps minimize crosstalk, avoid breakdowns, minimize leakage currents, ensure more homogeneous heat dissipation, etc. Circuit Sizing [ 231 ]: Circuit area minimization is critical for practically all microelectronic devices and systems, especially for implantable and wearable healthcare devices and generally those where the area of the circuit is limited by design requirements. Optimal circuit sizing is actually of interest for basically all microelectronic circuits since it helps improve overall performance and enhance circuit reliability.…”

“…In a similar work by Benhala [ 250 ] dedicated to the design of a two-stage op-amp with a minimized area and a maximized common mode rejection ratio, the authors concluded that the use of the ant colony optimization method shows a significantly lower convergence speed and increased computing time in comparison to their proposed combination of the backtracking search and ACO (BAACO) for the design of a two-stage op-amp with a minimized area and a maximized CMRR. An exemplary flowchart for a particular complete circuit design of a self-biased amplifier using metaheuristic methods is shown in Figure 28 [ 231 ].…”

“… Flowchart of a particular application of metaheuristic approach in analog circuit optimization [ 231 ]. …”