實現感應電動車最大加速性能與最高效率之控制設計與切換策略

Abstract

隨著近年來電動車技術以提高效率與扭力性能為目標發展，採用永磁馬達的高功率密度與高扭力密度的表現固然較佳，但感應馬達低成本、高耐用度與廣域操作速度的特性，為永磁馬達所不及之優勢；且在輕型電動車輛之應用上，更能突顯小功率感應馬達在高過負載能力的優勢。但欲達成感應馬達車用動力之高扭力與高效率的訴求，在控制技術上仍屬一大挑戰。因此，本研究致力於發展感應馬達之精密伺服控制，成功以德州儀器TMS320F28335 DSP晶片實現電動車兼具最大扭力與最高效率之自動切換控制策略。 在間接向量控制的實現中，為了達成磁場與扭力的動態控制，需進行感應馬達同步電氣角的估測，其計算乃建立在轉子時間常數(rotor time constant)的基礎上，此參數量測準確與否將影響控制的性能。有鑑於傳統轉子時間常數的量測方式實現困難，本研究發展以加速度為判斷依據的鑑別方法，不僅量測結果具有高準確度，且可直接在車輛上量測而更加容易實現；此外，為了達到單位電流最大扭力(maximum torque per amperage, MTPA)的控制特性，通常是將磁場電流(flux-producing current)與扭力電流(torque-producing current)設置於相同大小，但此方法在電動車的應用上造成了嚴重的磁飽和現象，使扭力輸出不如預期。本研究建立一套實驗標準程序，建立磁場操作電流與車輛加速度的關係，並選用額定激磁操作達成車輛的最大加速性能。 雖然MTPA擁有最大的加速表現，但在等速操作時效率較差。普遍的做法是在車輛加速過程後，藉由調整磁場電流達成效率的提升，但在電流切換時將造成車輛駕駛不平順的現象。本研究提出一套新式的切換控制策略，所定義之滑差因素(slip factor)將根據駕駛情況自動調整，在車輛低速時保有最大加速性能、在高速的情況則有最高的效率輸出，且實驗結果證明此方法在切換過程中具有更平順的電流操作；相較將磁場電流與扭力電流設為相同的方法，本控制策略在加速性、效率與車輛尾速部分可分別改善67%、273.6%與111.8%。因此，本研究成功實現了更適合電動車控制、且兼具高扭力效能及低耗電效率的感應馬達控制策略。

As the environmental crises become more significant, electric vehicles have drawn more attentions and have been rapidly developed in recent years. Even though permenant magnetic motors (PM) are usually preferred due to their inherent characteritics of high torque and power density, induction motors (IM) behave better on the aspects of cost, reliability, and with a wider operational speed range. Particularly, IMs dominate PMs with higher overload capability on applications of low-powered electric vehicle. However, control techniques of IMs are still chanllenging to meet high-torque and high-efficiency vehicle performance. Therefore, this study is dedicated to developing a DSP-based precise servo control with maximizing torque and efficiency for the IM implemented on electric vehicles, and the developed automatic control strategy is successfully implemented on the TMS320F28335 DSP microcontroller. The indirect vector control is commonly exploited for high performance servo control of the IM. To achieve control of the flux and torque, precise synchronous angle estimation based on the rotor time constant is crucial for stator current decoupling, and thus control performance is highly dependent on the accuracy of parameters. Due to the fact that conventional identification approaches are suffered from difficulty of implementation, an acceleration-based identification process which can be directly performed on the vehicle with satisfactory accuracy is developed in this thesis. In addition, the maximum torque per amperage (MTPA) operation is considerably important to provide desirable acceleration of the vehicle. It has been already recognized that by dividing the stator current equally into the flux-producing component and the torque-producing component, the MTPA can be thus obtained. Nevertheless, such arrangement causes severe flux saturation and the desirable operation is usually lost. In this Thesis, a standard experimental process is developed to find proper setting of the flux-producing current so that the maximum acceleration can be fulfilled. Despite of the maximum acceleration performance of the electric vehicle obtained by the MTPA operation, its operating efficiency is still comparatively low. In general, the efficiency can be improved by adjusting the flux-producing current. A novel switching control strategy where the defined slip factor can be automatically adjusted according to driving conditions is developed for achieving both the maximum acceleration and the highest efficiency, and experimental results show that it is more suitable to traction control of the electric vehicle because of its smooth operating performance. It has been also verified that all the acceleration, efficiency coefficient, and final vehicle speed can be improved by 67%, 273.6%, and 111.8%, respectively, when the system is fed by the maximum stator current compared with those of the conventional control approach by applying equal d-q current command.