Particle Swarm Optimization Based Design Optimization for Enhanced Breakdown Voltage in GaN High Electron Mobility Transistors

The dataset used for Particle Swarm Optimization (PSO)-based design optimization of GaN High Electron Mobility Transistors (HEMTs) consists of simulated device parameters and corresponding electrical performance metrics. The data is generated through Technology Computer-Aided Design (TCAD) simulations or extracted from experimentally validated models of GaN HEMTs.

Input Parameters (Design Variables)

These represent the structural and material properties of the device that are optimized using PSO:

Gate length (Lg)
Gate-to-drain distance (Lgd)
Barrier layer thickness
AlGaN composition (Al mole fraction)
Doping concentration
Passivation layer thickness
Field plate dimensions (if applicable)

These parameters directly influence the electric field distribution and breakdown characteristics of the device.

Output Parameters (Performance Metrics)

These are the target values evaluated for optimization:

Breakdown voltage (Vbr) (primary objective)
Drain current (Id)
Threshold voltage (Vth)
Electric field peak distribution
Leakage current
Nature of Data
The dataset is numerical and continuous
Each row corresponds to a unique device configuration
Data is typically nonlinear and high-dimensional, making it suitable for optimization using metaheuristic algorithms like PSO
Role in PSO Optimization
The dataset acts as the fitness evaluation base
PSO iteratively adjusts input parameters to maximize breakdown voltage while maintaining acceptable electrical performance
Fitness function is defined primarily based on maximizing Vbr with constraints on other parameters