Generation of shapers with neural networks to minimize noise in detection chains
Generation of shapers with neural networks to minimize noise in detection chains
Regadío , Dr. Alberto G. Tejedor, Dr. J. Ignacio 1Juan, Dr. J. Blanco
作者信息
- 1. Universidad de Alcalá Escuela Politécnica Superior
- 折叠
摘要
In this article, we present a method that uses an Artificial Neural Network (ANN) to maximize the Signal-to-Noise Ratio (SNR) of pulses from particle detector chains. However, rather than directly processing detector signals, the ANN is trained to produce the coefficients of a Finite Impulse Response (FIR) filter that maximize pulse height while minimizing noise. One advantage of this approach is that the training process requires only binary classification of input signals, distinguishing between signals containing particle-induced pulses and those without. Another is that the ANN is not required in the final system, just the hardware or software implementation of the FIR filter whose coefficients are produced by the method, significantly reducing design complexity. The simulation results demonstrate that this method reproduces established pulse-shaping filters for common noise characteristics. The approach can be tailored to specific detection chains, and to validate this methodology, we applied it to the Castilla-La Mancha (CaLMa) neutron monitor, achieving results consistent with theoretical expectations.
Abstract
In this article, we present a method that uses an Artificial Neural Network (ANN) to maximize the Signal-to-Noise Ratio (SNR) of pulses from particle detector chains. However, rather than directly processing detector signals, the ANN is trained to produce the coefficients of a Finite Impulse Response (FIR) filter that maximize pulse height while minimizing noise. One advantage of this approach is that the training process requires only binary classification of input signals, distinguishing between signals containing particle-induced pulses and those without. Another is that the ANN is not required in the final system, just the hardware or software implementation of the FIR filter whose coefficients are produced by the method, significantly reducing design complexity. The simulation results demonstrate that this method reproduces established pulse-shaping filters for common noise characteristics. The approach can be tailored to specific detection chains, and to validate this methodology, we applied it to the Castilla-La Mancha (CaLMa) neutron monitor, achieving results consistent with theoretical expectations.关键词
Digital pulse processing/Instrumentation and control (I&C)/Neural Networks/Digital shapers/Noise mitigation/Pulse height引用本文复制引用
Regadío , Dr. Alberto,G. Tejedor, Dr. J. Ignacio,Juan, Dr. J. Blanco.Generation of shapers with neural networks to minimize noise in detection chains[EB/OL].(2026-09-22)[2026-09-26].https://chinaxiv.org/abs/202609.00294.学科分类
电子技术概论
