Reinforcement Learning Electrical Engineering

Meng-Huan Chiang

Contact: *************@*****.*** 929-***-****

Website: LinkedIn GitHub

EDUCATION

Texas A&M University (TAMU) College Station, Texas M.S. in Electrical Engineering Aug. 2024 – Present National Cheng Kung University (NCKU) Tainan, Taiwan M.S. in Aeronautics and Astronautics Engineering Sep. 2022 – Jun 2024 B.S. in Aeronautics and Astronautics Engineering Sep. 2018 – Jun 2022 SKILLS

Python (NumPy, PyTorch, Gymnasium, Pandas, Scikit-learn, Matplotlib) MATLAB/Simulink (dynamics simulation, plot, animation) PUBLICATION

Autonomous Gain Tuning for Differential Drive Robots Targeting Control using Soft Actor-Critic, IEEE Conference on Artificial Intelligence (IEEE CAI 2024)

Resolved deadlock loop problems of the Approximate Pose Increment Control (APIC) law on Differential Drive Robot to achieve 100% success rate in targeting control tasks.

Integrating APIC with Reinforcement learning Soft Actor-Critic (SAC) agent. EXPERIENCE

Research Assistant Sep. 2022 – June 2024

Intelligent Embedded Control Lab, NCKU

Collaborated with the National Chung Shan Institute of Science and Technology in a 15-person team to conduct research on the application of Reinforcement Learning on unmanned aircraft.

Assisted with the annual project review by contributing to proposal writing and research reports. Teaching Assistant Sep. 2022 & June 2023

Department of Aeronautics and Astronautics Engineering, NCKU

Guided over 70 students with assignments and exams in Program Design (C programming) and Measurement and Signal Processing courses.

PROJECTS

Traffic Sign Recognition on CCTSDB dataset

Achieved 0.93 mAP50 in the task with 0.93 million of parameters by combining YOLOv5s with MobileNetV3 and Efficient Channel Attention module. Reinforcement Learning-based Targeting Control of Unmanned Aircraft

Achieved a 98% success rate in the pursuing tasks by training SAC agents with PyTorch.

Built a training environment with high-fidelity flight dynamics using Gymnasium and JSBSim. Autonomous Driving System Development for Differential Drive Robots

Led a 4-person team to develop an auto driving system for JetBot in Duckie Town environment.

Integrated and developed over 5 sub-systems using Python, including image preprocessing, lane detection, lane following with a PID controller, obstacle detection and path planning. Reinforcement Learning based Autonomous driving system

Trained a RL agent that is capable to drive along the road and avoid the obstacles based on the image input. Tainan City Rental Market Price Prediction

Ranked top 8 out of 70 participants by conducting data visualization, data preprocessing, data analysis, and regression modeling using pandas and scikit-learn. 2020 Drones Innovation and Application Competition

Won the High Distinction Award (Top 5 in Taiwan) for designing a tiltrotor-based VTOL unmanned aircraft.

Contact this candidate