职位描述
1.负责全局/局部路径规划算法(如A*、Dijkstra、RRT/RRT*、DWA、Lattice、Hybrid A*、APF等)的设计、实现与优化,满足动态环境下的实时性与鲁棒性要求; 2.解决复杂场景下的路径规划问题(如狭窄通道避障、动态障碍物预测、多机器人协同规划、三维空间规划等)。探索前沿技术,如基于深度学习(CNN/GNN)、强化学习的端到端规划算法开发; 3.优化路径的平滑性、安全性、能耗效率,支持多目标约束(时间最短、能耗最低、风险最小等); 4.与感知、控制、仿真团队协作,完成规划算法与传感器数据(激光雷达、摄像头、IMU)的融合调试; 5.低速场景行为决策、交通规则嵌入(如换道、路口通行)、紧急避障策略开发,人机协作场景安全规划、多机协同路径分配。
任职要求
1.硕士及以上学历(计算机、自动化、航空航天、测绘等专业),本科需具备3年以上相关项目经验; 2.精通C++/Python,熟悉Linux开发环境;掌握ROS/ROS2、PCL、OpenCV等开源框架; 3.深入理解A*、RRT系列、DWA等原理,具备算法改进经验(如RRT-Connect、Informed RRT*)。熟悉轨迹优化方法(多项式插值、样条曲线、最优控制),了解运动学/动力学约束下的可行性处理; 4.有实际路径规划项目落地经验(如自动驾驶决策模块、无人机航迹规划、仓储机器人调度系统); 5.理解感知(目标检测、跟踪)、控制(轨迹跟踪)模块的接口需求,具备系统集成调试能力; 6.掌握强化学习(PPO、DQN)在规划中的应用,或具备CUDA加速算法开发经验优先。
联系方式
上班地址:合肥, 中国
公司信息
SUNPURE TECHNOLOGY was founded in 2019. As a high-tech innovative enterprise, specializing in R & D, Manufacturing, Sales and Service of Intelligent PV Cleaning Robots, Sunpure Technology is committed to provide a standardized, professional and intelligent cleaning solution for solar power plants. Our R&D team is equipped with a group of professionals comprising of industry experts and research doctorate’s who have developed patented cloud-based IOT system. Up to now, the company has applied for 182+ patents, and participated in drafting the industry\'s first product Water-free Intelligent PV cleaning robot. The company takes customer achievement as the purpose of service, with professional products and complete global service system,to be the leading player in intelligent PV cleaning technology, improve more power generation income for photovoltaic power stations, and create greater asset value for customers.