Scholarlink

<p>Smart meter integrated circuits (ICs) within the VLSI system are designed to efficiently measure and manage electrical energy consumption in smart grid applications, relying on robust Power Distribution Networks (PDNs) for stable operation. However, sudden changes in load or switching activity in PDN delay the low-noise supply voltage in VLSI systems. Hence, a novel "Recurrent Adaptive Wavelet-Enhanced Noise Removal (RAWEN) with Synchronized Major-Minor Voltage Regulator" is proposed. Additionally, ground bounce due to transient currents disrupts noise prediction models, causing fluctuating ground potentials and signal distortion, which in turn affects the quality of the supply signal. Thus, the “Adaptive Moving Garrote-Wavelet Filtering Network” is incorporated within an LSTM layer, which incorporates the Adaptive Simple Moving Average (ASMA) algorithm to capture current consumption pattern using a dynamically adjustable window size and Fast Fourier Transform (FFT) to identify genuine current trends and ground noise-induced variations with Garrote Ricker wavelet filter thresholding algorithm to isolate and eliminate noise components from the FFT spectrum while maintaining current trends. Furthermore, Decoupling capacitors in voltage regulation methods help stabilize voltage levels in PDNs, but high-frequency switching operations cause timing errors and synchronization issues, reducing their effectiveness. So, Synchro-Proxi-Cap Major-Minor Voltage Regulator (SPCMM-VR) is introduced, which integrates Distributed Proxi Parallel Decoupling Capacitors to maintain consistent voltage levels across the chip during both normal and transient operating conditions with Dynamic Voltage and Frequency Scaling (DVFS) technique which enables dynamic voltage and frequency adjustments to reduce clock skews and synchronization issues during high-frequency transient conditions. Experimental results confirm the proposed model ensures low-noise supply voltage to ICs, improving voltage regulation during high-frequency switching and near-future noise prediction.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Recurrent adaptive wavelet-enhanced noise removal with synchronized major-minor voltage regulator for smart grid meter IC

  • 期刊论文

  • 发表日期：

    2025年12月18日

查看全文
  • C K Arunlal,
  • A Selwin Mich Priyadharson

摘要

Smart meter integrated circuits (ICs) within the VLSI system are designed to efficiently measure and manage electrical energy consumption in smart grid applications, relying on robust Power Distribution Networks (PDNs) for stable operation. However, sudden changes in load or switching activity in PDN delay the low-noise supply voltage in VLSI systems. Hence, a novel "Recurrent Adaptive Wavelet-Enhanced Noise Removal (RAWEN) with Synchronized Major-Minor Voltage Regulator" is proposed. Additionally, ground bounce due to transient currents disrupts noise prediction models, causing fluctuating ground potentials and signal distortion, which in turn affects the quality of the supply signal. Thus, the “Adaptive Moving Garrote-Wavelet Filtering Network” is incorporated within an LSTM layer, which incorporates the Adaptive Simple Moving Average (ASMA) algorithm to capture current consumption pattern using a dynamically adjustable window size and Fast Fourier Transform (FFT) to identify genuine current trends and ground noise-induced variations with Garrote Ricker wavelet filter thresholding algorithm to isolate and eliminate noise components from the FFT spectrum while maintaining current trends. Furthermore, Decoupling capacitors in voltage regulation methods help stabilize voltage levels in PDNs, but high-frequency switching operations cause timing errors and synchronization issues, reducing their effectiveness. So, Synchro-Proxi-Cap Major-Minor Voltage Regulator (SPCMM-VR) is introduced, which integrates Distributed Proxi Parallel Decoupling Capacitors to maintain consistent voltage levels across the chip during both normal and transient operating conditions with Dynamic Voltage and Frequency Scaling (DVFS) technique which enables dynamic voltage and frequency adjustments to reduce clock skews and synchronization issues during high-frequency transient conditions. Experimental results confirm the proposed model ensures low-noise supply voltage to ICs, improving voltage regulation during high-frequency switching and near-future noise prediction.