[Skip to main content](#main-content)

[OpenNash Jobs](https://jobs.opennash.com/)
[Hiring report](/reports/hiring/) · [All Zhongji Innolight jobs](https://jobs.opennash.com/companies/zhongji-innolight/) · [Search jobs](/search/)

Zhongji Innolight · 成都

# 算法类（研发）

成都 

Listed on Zhongji Innolight’s official careers site. Posted August 5, 2026. Check that it is still open and apply there.
[Apply on Zhongji Innolight’s careers site →](https://zj-innolight.jobs.feishu.cn/300308/position/7670385900632082742/detail)

[OpenNash Jobs](https://jobs.opennash.com/) / [Zhongji Innolight](https://jobs.opennash.com/companies/zhongji-innolight/) / 算法类（研发）

## Job description
- 负责机械臂/复合机器人的真机强化学习算法设计，融合视觉感知（目标/位姿）作为状态输入，构建从仿真训练到真机部署的完整流程，解决Sim2Real迁移问题
- 研发高效训练方案，优化样本效率，确保灵巧操作任务在有限时间内收敛到高成功率
- 搭建人机协作训练系统，利用遥操作、奖励设计、渐进式自主等策略降低真机训练成本 1. 2027届硕士及以上学历，计算机类、人工智能类、自动化类、电子信息类等相关专业 2. 211-双一流院校及以上优先 3. 熟悉协作机械臂的实际操作与底层接口，掌握正逆运动学求解及轨迹规划知识
- 具备机械臂/复合机器人真机强化学习项目经验，熟悉从仿真训练到真机策略部署的全流程，了解Sim2Real迁移的关键技术（域随机化、阻抗控制、自动重置等）
- 精通主流RL算法（SAC/PPO/TD3等），有面向机械臂灵巧操作的高样本效率训练实际调优经验
- 熟练使用PyTorch，具备并行仿真环境的搭建经验
- 良好的工程能力，熟悉Linux开发环境、ROS通信、GPU训练调度；具备计算机视觉基础（图像/点云处理、目标检测等），能将视觉信息与RL策略结合者加分
- 英语CET-6及以上

## At a glance
- Posted by the employer: Aug 5, 2026
- First seen by OpenNash: Oct 20, 2026
- Last checked: Oct 20, 2026
- Listed for: 76 days
- Location: 成都[Apply on Zhongji Innolight’s careers site →](https://zj-innolight.jobs.feishu.cn/300308/position/7670385900632082742/detail)

## More jobs at Zhongji Innolight

### [IE设计](https://jobs.opennash.com/companies/zhongji-innolight/jobs/d8829d82ed12f7afe39d/)

成都

### [IT系统方案工程师](https://jobs.opennash.com/companies/zhongji-innolight/jobs/0dbf27c024e56e3f67ca/)

成都

### [NPI类（新产品导入）](https://jobs.opennash.com/companies/zhongji-innolight/jobs/50a445bc82862f189cee/)

成都

### [PCB技术类（研发）](https://jobs.opennash.com/companies/zhongji-innolight/jobs/b00fa855a52a02fa92d9/)

成都

### [光学封装类（智能制造）](https://jobs.opennash.com/companies/zhongji-innolight/jobs/438f4851c3783b704d9d/)

成都

## Similar jobs at other companies

### [问界整车研发-整车标定&软件、算法类【2027届校招】](https://jobs.opennash.com/companies/seres-group/jobs/4317ba645ecb9a16fa27/)

Seres Group · 重庆市

OpenNash lists jobs from employers’ own careers sites and does not handle applications.
Dates come from the employer where stated; “first seen” and “last checked” are our own observations.

OpenNash Jobs · Official listings with dates and links to the employer. [Weekly hiring report](/reports/hiring/) · [Methodology](/methodology/)