Optimizing a GaN HEMT Model with ANN and PSO
Author Information
Author(s): Shen Haowen, Zhou Wenyong, Wang Jinye, Jin Hangjiang, Wu Yifan, Wang Junchao, Liu Jun, Malcovati Piero
Primary Institution: Innovation Center for Electronic Design Automation Technology, Hangzhou Dianzi University, Hangzhou, China
Hypothesis
Can integrating artificial neural networks with particle swarm optimization improve the parameter optimization of GaN HEMT models?
Conclusion
The ANN-PSO method significantly enhances the automation and efficiency of parameter optimization for GaN HEMT models while maintaining accuracy.
Supporting Evidence
- The ANN model achieved a simulation prediction accuracy of 99.9% for S-parameters.
- The optimized model showed an average error of 4.43% compared to measured data.
- The ANN-PSO method outperformed traditional optimization techniques in terms of speed and accuracy.
- Parameter optimization using ANN and PSO can significantly reduce the time required for model tuning.
Takeaway
This study shows how using a computer program can help make better models of electronic devices faster and easier.
Methodology
The study used an artificial neural network to predict S-parameters and a particle swarm optimization algorithm to optimize model parameters.
Limitations
The optimization method may not fully account for all variations in device behavior under different conditions.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website