114/3/6（四）國立中正大學機械工程學系 黃旭志 教授
- 演講時間：114/3/6（四）15:30-17:00
- 演講地點：國立清華大學工程一館107演講廳
- 演講者： 黃旭志教授 (國立中正大學 機械工程學系)
- 講題：使用穩健正交螢火蟲演算法對群體機器人資訊網宇實體系統進行無碰撞模糊編隊控制 (Collision-Free Fuzzy Formation Control of Swarm Robotic Cyber-Physical Systems Using a Robust Orthogonal Firefly Algorithm)
- Abstract:
This talk presents a collision-free fuzzy formation control strategy of swarm robotic cyber-physical systems (CPS) using a robust orthogonal firefly algorithm (OFA). The classical FA is fused with the Taguchi method and ranking mutation process to present a hybrid artificial intelligence. This computational intelligence is employed with fuzzy theory to develop optimal cyber and cognition levels in swarm robotic CPS 5C architecture. With the robotic sensors and actuators, the OFA-fuzzy based cyber and cognition levels are incorporated with smart connection, data-to-information conversion and configuration levels to design a pragmatic swarm robotic CPS using system-on-a-programmable-chip (SoPC) technology. The broadcast distributed control strategy and potential field are employed to address the formation control problem of swarm robotic CPSs with obstacles. This approach avoids the communication delay in conventional graph theory based consensus multi-robot CPSs under a directed network topology. In the proposed swarm robotic CPS, the embedded central processing unit (CPU), operating system (OS), networking intellectual property (IP) and robotic custom IPs are integrated in field-programmable gate array (FPGA) chips. The proposed SoPC-based OFA-fuzzy swarm robotic CPS features the seamless integration of computation, networking, and physical robotic process. Experimental results and comparative works are conducted to show the merits of the proposed swarm robotic CPS to achieve collision-free distributed formation control.