FPGA Implementation of a Multi-Layer Perceptron
2025 International Conference on Electrical Drives and Power Electronics (EDPE)
Resumen
Artificial neural networks are defined as nonlinear mapping systems whose structure is based on principles observed in human and animal nervous systems. It is an artificial neural network widely used in classification and regression problems. This paper presents a methodology for implementing an artificial intelligence (AI)-based control algorithm for a buck converter using field-programmable gate arrays (FPGAs). The proposed approach leverages the error between the reference voltage and the output voltage as the primary input, and the compensator output as the target, enabling the neural network to learn the control behavior offline and subsequently replicate it in real time. The FPGA platform was selected due to its ability to execute highly parallel operations, offering significant improvements in processing speed, customization, and power efficiency compared to conventional platforms such as GPUs. Nevertheless, challenges arise in hardware design and algorithm adaptation, as FPGA implementation requires careful consideration of numerical precision, resource allocation, and hardware description languages. The results demonstrate that FPGA-based AI controllers are a promising alternative for power electronic converters, combining high performance with adaptability.
Datos de publicación
Cita esta publicación
Comprueba el estilo que requiere tu institución o revista.
Fuentes y trazabilidad
Se conserva la procedencia de los datos y se da prioridad a la revista cuando existe información editorial.
Licencias declaradas por las fuentes
https://doi.org/10.15223/policy-029
https://doi.org/10.15223/policy-037