<p>An enhanced parameter-free predictive current control is proposed to address the shortcomings of conventional predictive current control, including weak parameter robustness and significant current ripple. It is implemented within a three-level inverter for driving permanent magnet synchronous motor (PMSM) systems. Initially, by virtue of a full-current-variation update technique, the future current can be predicted to facilitate the first applied vector, mitigating the impact of parameter mismatch. Subsequently, a current variations correction technique is employed to refine the predictive accuracy of future current, and the optimal duration is decided via the dead-beat principle. Furthermore, the second active vector is selected by an improved cost function, which incorporates considerations such as the pre-selection of voltage vectors and the neutral-point voltage constraint. Finally, experimental results have demonstrated the superiority of the proposed strategy in current ripple suppression along with its strong robustness against parameter mismatch.</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

Low current ripple parameter-free predictive current control of three-level inverter-fed PMSMs with current variations correction

  • Zinuo Wang,
  • Feng Yu,
  • Jin Zhang,
  • Xiaopeng Qin

摘要

An enhanced parameter-free predictive current control is proposed to address the shortcomings of conventional predictive current control, including weak parameter robustness and significant current ripple. It is implemented within a three-level inverter for driving permanent magnet synchronous motor (PMSM) systems. Initially, by virtue of a full-current-variation update technique, the future current can be predicted to facilitate the first applied vector, mitigating the impact of parameter mismatch. Subsequently, a current variations correction technique is employed to refine the predictive accuracy of future current, and the optimal duration is decided via the dead-beat principle. Furthermore, the second active vector is selected by an improved cost function, which incorporates considerations such as the pre-selection of voltage vectors and the neutral-point voltage constraint. Finally, experimental results have demonstrated the superiority of the proposed strategy in current ripple suppression along with its strong robustness against parameter mismatch.