題目：New Insights into Autonomous Unmanned Vehicles and Embedded Systems: Convex Optimization and Soft Computing Methods

報告人：Chaomin Luo教授，University of Detroit Mercy, Michigan, USA

時間：2015.6.25, 9:30

地點：信息工程學院329室

Chaomin Luo（雒超民），本科畢業于東南大學無線電工程系，2002年在加拿大圭爾夫（Guelph）大學獲得電子工程專業碩士學位。2008年在加拿大滑鐵盧大學（Waterloo）電氣與計算機工程系獲得博士學位。2008年博士畢業后在國立臺北大學電機工程研究所任助理教授及電機與資訊工程學院院長助理，目前為美國底特律大學電氣與計算機工程系高級移動機器人實驗室任終身教授。

主要從事人工智能及其應用，嵌入式系統，智能機器人，信息融合及超大規模集成電路的優化設計等方面的研究。在世界上首次成功將半正定規劃，凸規劃和基于二階錐規劃等優化算法用于超大規模集成電路設計，也是國際上首次采用基于生物激勵的神經網絡算法和模型用于智能機器人的研究者（如智能清掃移動機器人和智能探礦機器人）。

歷任IEEE美國東南密歇根人工智能分會主席，副主席等職。雒博士現任IEEE美國西南密歇根州教育專家委員會主席，IEEE美國東南密歇根人工智能分會主席，和IEEE美國東南密歇根機器人和自動化分會副主席等學術職位。國際雜志《International Journal of Robotics and Automation》等4個期刊編委。報告內容簡介：

Nowadays, optimization and soft computing methodologies play increasingly important role on electrical engineering. In this research, convex optimization and soft computing methods are applied for VLSI design and intelligent vehicle navigation, control and mapping.

A sequence of novel neural dynamics, genetic algorithms (GAs), and fuzzy logic (FL) approaches associated with developed, numerical method, spline-based and vector-driven intelligent vehicle navigation model are proposed. The biologically-motivated neural networks (BNN) algorithms are employed to guide an intelligent vehicle to reach goal with obstacle avoidance motivated by a biological neural system. A spline-based smooth guidance paradigm is developed for guidance of the vehicle locally so as to plan more reasonable and smoother trajectories with GAs and FL methodologies. BNN based scheme demonstrates that the algorithms avoid the issue of local minima in path planning of an intelligent vehicle. Simulation, comparison studies and experimental results of intelligent vehicle navigation demonstrate the effectiveness, efficiency and robustness of the proposed methodologies.