Reinforcement learning

<p>Duties:</p><p>1. Responsible for collecting, simulating, training, and deploying human-shaped robot motion control data. </p><p>2. Track current robot-related reinforcement learning, system identification, etc. frontier technologies, and combine existing algorithms for optimization.</p><p>3. Identify system performance bottlenecks, propose validation and improvement solutions, and continuously optimize algorithm stability and efficiency.</p><p>4. Write algorithm design documents and application deployment scripts.</p><p>Requirements:</p><p>1. Computer science, robotics, artificial intelligence, engineering, mechanics, etc. related fields. </p><p>2. Have reinforcement learning, deep learning related theory and application background. Familiar with Gazebo, MuJoCo, isaac, etc. Physical simulation application.</p><p>3.3 years of experience in humanoid robot bipedal walking/dual-arm control, with an understanding of the system structure of multi-degree-of-freedom robots and mechanical arm control algorithms.</p><p>4. Proficient in Python/C/C++ languages, familiar with Linux and ROS environment usage, familiar with Git, Docker, etc. Tools. </p><p>5. Have good team cooperation and communication skills.</p>