電動車混合電源之隨機自優化功率管理系統

Stochastic Self-Optimizing Power Management for Hybrid Power Sources of Electric Vehicles

李昭明(Chao-Ming Lee)

指導教授 ： 林巍聳

國立臺灣大學/電機資訊學院/電機工程學研究所/博士(2015年)

https://doi.org/10.6342/NTU.2015.01630

摘要

本論文探討電動車混合電源的功率管理策略，第一種混合電源由鋰離子電池與超級電容組合而成，透過一個雙向升降壓直流對直流轉換器控制電流，混合電源能供電給負載也能回收再生電能；第二種混合電源由燃料電池與被動混合儲能系統組合而成，透過一個單向升壓直流對直流轉換器控制燃料電池的電流。針對這兩種混合電源系統，我們分別提出隨機功率管理策略與隨機自優化功率管理策略，這兩種功率管理策略都用輻射基底函數類神經網路建構，然後利用強化學習架構與極小值原理進行優化，目標是在隨機行車行程之下，功率管理策略能夠決定最佳的主電源輸出功率。本論文採用馬可夫鏈建構隨機行車行程，特點是可以充分涵蓋實際駕駛狀況，因此對隨機行車行程優化的功率管理策略可以處理實際的行車狀況；在鋰離子電池與超級電容的混合電源架構中，隨機功率管理策略控制鋰離子電池的輸出功率，並且利用超級電容供應瞬間高功率以及吸收大部份的再生電能，同時維持足夠的超級電容殘餘能量。在燃料電池與被動混合儲能系統的混合電源架構中，隨機自優化功率管理策略控制燃料電池的輸出功率，使燃料電池系統的發電效率最大化，並且利用被動式儲能系統供應瞬間高功率以及吸收大部份的再生電能，同時維持適當的鋰離子電池殘電量，並且保護鋰離子電池不受損壞；模擬結果顯示，不論是隨機功率管理策略或是隨機自優化功率管理策略，都能夠在標準行車行程和隨機行車行程獲得優良的功率管理成效。

關鍵字

混合電源系統 ； 隨機行車行程 ； 電動車 ； 自優化功率管理策略

並列摘要

This dissertation concerns about the power management strategy of electric vehicles supplied by hybrid power sources (HPS). The first type of the HPS is composed of a Lithium-ion battery (LIB) pack and an ultracapacitor (UC) bank. The LIB/UC HPS can provide load power and receive regenerative power by controlling the current through the bi-directional buck/boost DC/DC converter. The second type of the HPS consists of the fuel-cell system (FCS) and the passive hybrid energy storage system (HESS). For these HPSs, we proposed the stochastic power management strategy (SPMS) and the stochastic self-optimizing power management strategy (SSOPMS), respectively. Both these two strategies are composed of radial basis function neural networks (RBFNN), which are trained and optimized by the reinforcement learning scheme and the minimum principle. The main objective of these PMSs is to determine the optimal output power of the major power source through the stochastic driving cycles. In this dissertation, stochastic driving cycles are implemented by Markov chain that can take more complex and real-world driving conditions into account. Therefore, the PMSs optimized through the stochastic driving cycles can deal with driving conditions. For the LIB/UC HPS, the SPMS controls the output power of the LIB and utilizes the UC to provide transient peak load power and to receive most of the regenerative power. Meanwhile, enough residue energy of the UC bank is maintained. For the FCS and passive HESS HPS, the SSOPMS controls the output power and maximizes the operating efficiency of the FCS. Moreover, the SSOPMS utilizes the passive HESS to furnish transient peak load power and receive most of the regenerative power. Meanwhile, it can maintain proper residue charge of the LIB and keep the LIB from damage. The simulation results show that both the SPMS and the SSOPMS can achieve excellent power management on standard and stochastic driving cycles.

並列關鍵字

Hybrid power system ； stochastic driving cycle ； electric vehicle ； self-optimizing power management strategy

參考文獻

A. Arce, A. J. del Real and C. Bordons (2009). "MPC for battery/fuel cell hybrid vehicles including fuel cell dynamics and battery performance improvement." Journal of Process Control 19(8): 1289-1304.

D. P. Bertsekas (2005). Dynamic programming and optimal control, Athena Scientific Belmont, MA.

J. Cao and A. Emadi (2012). "A New Battery/UltraCapacitor Hybrid Energy Storage System for Electric, Hybrid, and Plug-In Hybrid Electric Vehicles." Power Electronics, IEEE Transactions on 27(1): 122-132.

R. Carter, A. Cruden and P. J. Hall (2012). "Optimizing for Efficiency or Battery Life in a Battery/Supercapacitor Electric Vehicle." Vehicular Technology, IEEE Transactions on 61(4): 1526-1533.

B.-C. Chen, Y.-Y. Wu, Y.-L. Wu and C.-C. Lin (2011). "Adaptive Power Split Control for a Hybrid Electric Scooter." Vehicular Technology, IEEE Transactions on 60(4): 1430-1437.