Yi Niu




Award Achievements

Project Experience

  1. China Robot Competition CSU_YUNLU team
    • Time : 2021.04-2022.04
    • Project Description: The content of this competition is the simulation of the RoboCup five-a-side robot football game. The simulation platform is the Gezebo platform built on ROS, relying on robocup to simulate the robot control algorithm.
    • Responsibilities:
      • As one of the main leaders of the school in this competition, he was responsible for the writing of the robot defense and avoidance and obstacle avoidance strategy codes for the competition (with the help of path planning based on sampling-based pathfinding algorithm.
      • Use the integral separation PID algorithm with dead zone to optimize movement and steering strategies, speeding up robot movement control).
      • Responsible for writing the relevant content of the algorithm section of the paper and participating in the implementation of the project algorithm. Currently, the paper has been published by OP and indexed by EI
      • Represent the team at the 2021ICRAIC conference and conduct workshops.
    • Project Achievements:
      • 2021 China Robot Competition Medium Simulation Group Champion (First Prize)
      • 2021 China Robot Competition Technical Challenge Champion
      • Published papers at the 2021ICRAIC International Conference
  2. UAV Swarm Assists Edge Computing Offloading Project Team
    • Time : 2022.03-2022.05
    • Project Description : This project is based on a small-scale large-scale data throughput scenario (such as nucleic acid detection). It envisions using drones to complete edge-assisted computing offloading tasks, and allocate computing power to the offloading tasks in two ways: within the cluster and between clusters.
    • Responsibilities:
      1. Plan the flight path of the UAV cluster when the ground target cluster moves to ensure that the ground target is tracked while the data transmission distance is as small as possible;
      2. The k-means algorithm using Thiessen polygon and unsupervised clustering is used to assign clusters to UAVs.
  3. Chinese Medical Information Processing CMeEE Competition Team
    • Time : 2022.03-2022.05
    • Project Description : This competition requires the development of a named entity recognition model based on the BERT pre-training model for medical named entity recognition, that is, extracting entity nouns related to clinical medicine from a large amount of medical text. The F1 value is used as the evaluation standard and the competition is conducted in the form of rankings.
    • Responsibilities: The optimization of network models and the tuning of model parameters.