摘要
本研究的目的在研製一套新型「液靜壓傳動變速系統」及建立其換檔控制策略。「液靜壓傳動變速系統」以一套液靜壓系統(含一對可變排量雙向液壓泵/液壓馬達及蓄壓器)和一個二自由度五桿行星齒輪組耦合形成一新型傳動系統來取代傳統之無段變速器(Continuously Variable Transmission, CVT)，以達成最佳之能源效率。由於液靜壓系統可快速且有效的回收大部份煞車應消耗能量，且於下次起動時協助動力馬達來加速車輛起動，與傳統變速系統相比，液靜壓傳動變速系統具有耗能低、加速性能佳、煞車保養減少、及電瓶壽險長等優點。 本研究，首先分析及評估液靜壓傳動系統之運轉方式及控制策略。並配合系統需求，設計、製作液靜壓傳動變速控制系統之測試平台，以進行驅動馬達及液靜壓傳動系統等次系統的性能測試與整體性能分析。在控制策略方面，本文提出一種以基因演算法為基礎的適應性類神經滑動模式控制器(GANSMC)，能以最有效率的方式找出最適合之適應控制参數，使變速系統能依據最佳耗能效率曲線，穩定且有效的進行換檔動作。實驗結果，與傳統PI控制比較，此適應控制系統有較佳的穩定性及強健性，可使換檔動作更平順穩定。
並列摘要
In this research, a new hydrostatic-transmission with shift control scheme is developed. Coupled with a 2-DOF planetary gear set, a hydraulic package of a reversible variable-displacement hydraulic pump/motor and accumulator is used to replace the conventional continuously variable transmission (CVT) to optimize energy economy. With this hydrostatic system, nearly all of the energy typically lost during vehicle braking is captured and used to propel the vehicle the next time it needs to accelerate. Compare with conventional CVT system, the benefits of hydrostatic-transmission system include: better energy economy, better acceleration, less brake maintenance, and longer battery life. In this study, the prime task is to analyze the operating characteristics and performance of the hydrostatic-transmission system and derive an optimal control strategy for power shift. Then, a test bench of hydrostatic-transmission system is arranged to perform operating tests for verifying the performances of the developed systems. For optimizing energy economy and smoothing shift control, a genetic algorithm (GA) based adaptive neural sliding mode control (GANSMC) is proposed to control the hydrostatic -transmission system. The control parameters of GANSMC can be properly adjusted by using online parameters tuning algorithm derived from GA. Compare with PI controller, the experimental results show that this novel controller has good control performance, stability, and robustness.