Projects

2020-2021: PID-Based Mecanum-wheel AGV Design

Type: Freshman Annual Project
Role: Leading
Grade: Excellent

Designed a cascaded PID controller for wheel speed and vehicle attitude control. Developed overall vehicle control programs based on microcontrollers, calculated AGV attitude through sensor information for decision-making and control. Combined with the omnidirectional movement characteristics of Mecanum wheels, achieved directional driving and autonomous obstacle avoidance for the AGV.

2021-2022: Deep Learning-Based Multi-Task Smart Vehicle Design

Type: Undergraduate Research Project
Role: Leading
Grade: Excellent

Utilized cameras to collect lane driving data and target images for annotation, training, and online deployment, enabling the smart vehicle to complete designated tasks such as lane keeping, target recognition, and parking. Designed mechanical structures to achieve point-to-point target grasping and placement, as well as striking specific targets.

Summer 2023: Intelligent Vehicle Platform Development

Type: Course Design Project
Role: Perception Module

Implemented lane detection and target recognition using cameras, performed point cloud preprocessing using PCL library, conducted real-time mapping and localization based on A-LOAM algorithm, and designed AEB (Autonomous Emergency Braking) algorithm based on filtered point cloud data to achieve autonomous emergency stopping.

2023-2024: Terrain-Complexity-Aware Multi-Mode Navigation System for Planetary Rovers

Type: Undergraduate Thesis
Role: Independent Project
Grade: Excellent

Developed a multi-mode autonomous navigation system for planetary rovers targeting planetary surface environments. Created a terrain complexity classifier based on planetary surface geometric features, developed adaptive perception, planning, and control algorithms for different terrain complexities, and designed a multi-mode navigation framework for autonomous navigation algorithm integration and adaptive mode switching.

2025-Present: Online Active Perception of Physical Properties for Planetary Rovers in Unknown Lunar Terrain

Type: Master’s Thesis
Role: Independent Project (Ongoing)

Research focuses on developing methods for planetary rovers to actively perceive and learn physical properties of unknown lunar terrain. The work aims to combine visual and tactile sensing for real-time terrain analysis and property estimation.