本文提出了一種針對內藏式永磁同步馬達電機驅動的強健無差拍電流控制，該方式建立在智慧型積分滑模控制基礎上，並結合自適應類神經網路。無差拍電流控制以其對電機驅動參數變化和外部干擾的敏感性而聞名，因此對其進行增加強健性的需求日益迫切。為了解決這個問題，本研究旨在減輕無差拍電流控制的參數敏感性，同時增強對干擾的整體強健性。首先推導考慮內藏式永磁同步馬達 dq軸時間延遲影響的無差拍電流控制建模和控制策略，還計算了dq軸的干擾項。提供積分滑模控制的詳細分析，該分析可以應對內藏式永磁同步馬達驅動的dq軸電流控制中的模型參數不匹配和干擾。此外為減小積分滑模控制的開關增益，提出使用自適應神經網路來估計dq軸干擾項的方法，從而實現智慧型積分滑模控制。實驗所使用之硬體為應用德州儀器公司生產之浮點數數位訊號處理器TMS320F28075之內藏式永磁同步馬達伺服驅動系統。;This study introduces a robust deadbeat predict current control (DPCC) scheme designed for an interior permanent magnet synchronous motor (IPMSM) drive. The scheme is built upon an intelligent integral sliding mode control (ISMC) by using an adaptive neural network (ANN). The DPCC, known for its sensitivity to motor drive parameter variations and external disturbances, prompted the need for enhanced robustness in current control. To address this, the proposed approach aims to mitigate the parameter sensitivity of DPCC and enhance the overall robustness of current control against disturbances. In this study, first, the modeling and control strategies of DPCC considering the effect of time delay for the dq-axis of IPMSM are derived. The disturbance terms of dq-axis are also formulated. Then, the detailed analyses of ISMC which can confront the model parameter mismatch and disturbances for the dq-axis current control of the IPMSM drive are provided. In addition, in order to reduce the switch gains of the ISMC, an ANN is proposed to estimate the disturbance terms of dq-axis resulted in an intelligent ISMC. Additionally, some experimental results are demonstrated to verify the effectiveness of the proposed robust DPCC using intelligent ISMC for the IPMSM drive in the constant toque region.
