Zhenchang Du , Shaoxiang Ma , Hongqi Zhang , Yixiang Huang , Ming Zhang
{"title":"利用有源滤波器抑制倒置型高压AGPS输出电压纹波","authors":"Zhenchang Du , Shaoxiang Ma , Hongqi Zhang , Yixiang Huang , Ming Zhang","doi":"10.1016/j.fusengdes.2025.115449","DOIUrl":null,"url":null,"abstract":"<div><div>This paper presents a DC active power filter (APF) designed to mitigate output voltage ripple in the acceleration grid power supply (AGPS), to enhance the operational efficiency and stability of negative ion-based neutral beam injection (N-NBI) systems. The proposed APF employs a 5-level active neutral-point-clamped (5L-ANPC) topology, integrated into the AGPS high-voltage filtering branch. An optimized model predictive control (MPC) strategy is developed, incorporating adaptive weighting coefficients and switching penalties to ensure DC-link capacitor voltage stabilization and balanced switching losses. Furthermore, a long short-term memory (LSTM) neural network is implemented to predict load current fluctuations, enabling the establishment of a comprehensive state-space model for the entire circuit. Simulation results demonstrate significant performance improvements: peak-to-peak ripple voltage is reduced to 31.66% of the original value (5689 V <span><math><mo>→</mo></math></span> 1601 V), while ripple voltage root mean square (RMS) is reduced to 22.56% of the original value (1764 V <span><math><mo>→</mo></math></span> 398 V).</div></div>","PeriodicalId":55133,"journal":{"name":"Fusion Engineering and Design","volume":"222 ","pages":"Article 115449"},"PeriodicalIF":2.0000,"publicationDate":"2025-09-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Output voltage ripple suppression in inverted type high-voltage AGPS using APF\",\"authors\":\"Zhenchang Du , Shaoxiang Ma , Hongqi Zhang , Yixiang Huang , Ming Zhang\",\"doi\":\"10.1016/j.fusengdes.2025.115449\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><div>This paper presents a DC active power filter (APF) designed to mitigate output voltage ripple in the acceleration grid power supply (AGPS), to enhance the operational efficiency and stability of negative ion-based neutral beam injection (N-NBI) systems. The proposed APF employs a 5-level active neutral-point-clamped (5L-ANPC) topology, integrated into the AGPS high-voltage filtering branch. An optimized model predictive control (MPC) strategy is developed, incorporating adaptive weighting coefficients and switching penalties to ensure DC-link capacitor voltage stabilization and balanced switching losses. Furthermore, a long short-term memory (LSTM) neural network is implemented to predict load current fluctuations, enabling the establishment of a comprehensive state-space model for the entire circuit. Simulation results demonstrate significant performance improvements: peak-to-peak ripple voltage is reduced to 31.66% of the original value (5689 V <span><math><mo>→</mo></math></span> 1601 V), while ripple voltage root mean square (RMS) is reduced to 22.56% of the original value (1764 V <span><math><mo>→</mo></math></span> 398 V).</div></div>\",\"PeriodicalId\":55133,\"journal\":{\"name\":\"Fusion Engineering and Design\",\"volume\":\"222 \",\"pages\":\"Article 115449\"},\"PeriodicalIF\":2.0000,\"publicationDate\":\"2025-09-20\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Fusion Engineering and Design\",\"FirstCategoryId\":\"5\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S0920379625006453\",\"RegionNum\":3,\"RegionCategory\":\"工程技术\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"NUCLEAR SCIENCE & TECHNOLOGY\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Fusion Engineering and Design","FirstCategoryId":"5","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0920379625006453","RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"NUCLEAR SCIENCE & TECHNOLOGY","Score":null,"Total":0}
Output voltage ripple suppression in inverted type high-voltage AGPS using APF
This paper presents a DC active power filter (APF) designed to mitigate output voltage ripple in the acceleration grid power supply (AGPS), to enhance the operational efficiency and stability of negative ion-based neutral beam injection (N-NBI) systems. The proposed APF employs a 5-level active neutral-point-clamped (5L-ANPC) topology, integrated into the AGPS high-voltage filtering branch. An optimized model predictive control (MPC) strategy is developed, incorporating adaptive weighting coefficients and switching penalties to ensure DC-link capacitor voltage stabilization and balanced switching losses. Furthermore, a long short-term memory (LSTM) neural network is implemented to predict load current fluctuations, enabling the establishment of a comprehensive state-space model for the entire circuit. Simulation results demonstrate significant performance improvements: peak-to-peak ripple voltage is reduced to 31.66% of the original value (5689 V 1601 V), while ripple voltage root mean square (RMS) is reduced to 22.56% of the original value (1764 V 398 V).
期刊介绍:
The journal accepts papers about experiments (both plasma and technology), theory, models, methods, and designs in areas relating to technology, engineering, and applied science aspects of magnetic and inertial fusion energy. Specific areas of interest include: MFE and IFE design studies for experiments and reactors; fusion nuclear technologies and materials, including blankets and shields; analysis of reactor plasmas; plasma heating, fuelling, and vacuum systems; drivers, targets, and special technologies for IFE, controls and diagnostics; fuel cycle analysis and tritium reprocessing and handling; operations and remote maintenance of reactors; safety, decommissioning, and waste management; economic and environmental analysis of components and systems.