Scholarlink

<p>Aircraft power grids often face challenges such as high harmonics and reactive currents caused by nonlinear onboard applications. To address these issues and ensure a more reliable and high-quality power supply, we propose a novel system called E-PATCH (Electric aircraft Power distribution system with Advanced Temporal CNN for Harmonic Cancellation), implemented through a Shunt Active Power Filter (SAPF). The proposed method employs a Temporal Convolutional Neural Network (TCNN) to forecast future current and voltage waveforms based on synchronized reference frame inputs to minimize switching losses. Moreover, a Fractional Order PID (FOPID) controller is utilized and the parameters of the controller are optimally adjusted by the Snow Leopard Optimization algorithm to provide a better performance of the system. The Voltage Source Inverter (VSI) is managed using Pulse Width Modulation (PWM) and injects compensating currents to eliminate harmonics, voltage dips, and flicker. Moreover, the balancing of voltages is made easy with the use of one DC link in proposed SAPF. The proposed method performance is evaluated under different loading conditions, with generator frequencies ranging from 400 to 800&#xa0;Hz, and tested using nonlinear loads with balanced grid voltages in MATLAB/SIMULINK. The proposed method reduces current THD from 24.8 to 3.8% and voltage THD from 26.5 to 4.9%, achieving compliance with IEEE-519 standards. The THD of the proposed controller 39.14%, 34.22%, and 12.2% lower compared to existing PI, PID, FOPID, and Fuzzy FOPID controllers, respectively.</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

E-PATCH: Enhancing Aircraft Power Grid Quality Through TCNN-Predicted Harmonic Cancellation

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年3月4日

查看全文
  • Sathyamoorthi Sethuram,
  • Selvaperumal Sundaramoorthy

摘要

Aircraft power grids often face challenges such as high harmonics and reactive currents caused by nonlinear onboard applications. To address these issues and ensure a more reliable and high-quality power supply, we propose a novel system called E-PATCH (Electric aircraft Power distribution system with Advanced Temporal CNN for Harmonic Cancellation), implemented through a Shunt Active Power Filter (SAPF). The proposed method employs a Temporal Convolutional Neural Network (TCNN) to forecast future current and voltage waveforms based on synchronized reference frame inputs to minimize switching losses. Moreover, a Fractional Order PID (FOPID) controller is utilized and the parameters of the controller are optimally adjusted by the Snow Leopard Optimization algorithm to provide a better performance of the system. The Voltage Source Inverter (VSI) is managed using Pulse Width Modulation (PWM) and injects compensating currents to eliminate harmonics, voltage dips, and flicker. Moreover, the balancing of voltages is made easy with the use of one DC link in proposed SAPF. The proposed method performance is evaluated under different loading conditions, with generator frequencies ranging from 400 to 800 Hz, and tested using nonlinear loads with balanced grid voltages in MATLAB/SIMULINK. The proposed method reduces current THD from 24.8 to 3.8% and voltage THD from 26.5 to 4.9%, achieving compliance with IEEE-519 standards. The THD of the proposed controller 39.14%, 34.22%, and 12.2% lower compared to existing PI, PID, FOPID, and Fuzzy FOPID controllers, respectively.