実環境における歩行安定性と移動効率を考慮した２足歩行ロボットの歩行軌道計画

考虑真实环境中行走稳定性和运动效率的双足机器人行走轨迹规划

• 批准号: 15J06497
• 负责人: ブランダン マルティン
• 金额: $ 1.47万
• 依托单位: Waseda University
• 依托单位国家: 日本
• 项目类别: Grant-in-Aid for JSPS Fellows
• 财政年份: 2015
• 资助国家: 日本
• 起止时间: 2015-04-24 至 2017-03-31
• 项目状态: 已结题
• 来源: https://kaken.nii.ac.jp/grant/KAKENHI-PROJECT-15J06497/
• 关键词: motion planning; footstep planning; humanoid robots; computer vision; friction from vision; material recognition; CNN; friction estimaton; stereo vision

1) I developed an algorithm to predict friction and its uncertainty from images at all pixels, based on material recognition.2) I improved the motion planning system, which now accounts for energy, stability, friction, collision, and uncertainty in perception. These are accounted for at both footstep and full-body motion planning levels thanks to a hierarchical architecture with oracles at lower levels predicting costs at higher levels.3) I evaluated the whole system on the real robot and in realistic simulations, in environments with different ground materials. The system was able to correctly identify the materials and friction, thus making intelligent footstep planning choices.The research was presented at the IEEE Humanoids conference and was a finalist for the Best Oral Paper award.

1)我开发了一种算法，基于材质识别，从所有像素的图像中预测摩擦及其不确定性。2)我改进了运动规划系统，该系统现在考虑了能量、稳定性、摩擦、碰撞和感知中的不确定性。这些都在足迹和全身运动规划级别上进行了考虑，这要归功于一个层次结构，其中较低级别的先知预测较高级别的成本。3)我在真实机器人上和在具有不同地面材料的环境中的真实模拟中对整个系统进行了评估。该系统能够正确识别材料和摩擦力，从而做出智能足迹规划选择。这项研究在IEEE类人会议上发表，并入围最佳口头论文奖。

项目成果

Footstep Planning for Slippery and Slanted Terrain Using Human-Inspired Models

• DOI: 10.1109/tro.2016.2581219
• 发表时间: 2016-07
• 期刊: IEEE Transactions on Robotics
• 影响因子: 7.8
• 作者: Martim Brandao;K. Hashimoto;J. Santos-Victor;A. Takanishi

On Stereo Confidence Measures for Global Methods: Evaluation, New Model and Integration into Occupancy Grids

• DOI: 10.1109/tpami.2015.2437381
• 发表时间: 2016
• 期刊: IEEE Transactions on Pattern Analysis and Machine Intelligence
• 影响因子: 23.6
• 作者: Martim Brandao;R. Ferreira;K. Hashimoto;A. Takanishi;J. Santos-Victor

Optimizing Energy Consumption and Preventing Slips at the Footstep Planning Level

在脚步规划层面优化能源消耗并防止失误

• 发表时间: 2015
• 作者: Martim Brandão;Kenji Hashimoto;José Santos-Victor and Atsuo Takanishi
• 通讯作者: José Santos-Victor and Atsuo Takanishi

Friction from Vision: A Study of Algorithmic and Human Performance with Consequences for Robot Perception and Teleoperation

视觉摩擦：算法和人类表现对机器人感知和远程操作影响的研究

• 发表时间: 2016
• 作者: Martim Brandao;Kenji Hashimoto;Atsuo Takanishi
• 通讯作者: Atsuo Takanishi

Extending humanoid footstep planning with ZMP tracking error constraints

使用 ZMP 跟踪误差约束扩展人形足迹规划

• 发表时间: 2015
• 作者: Martim Brandao;Kenji Hashimoto;Atsuo Takanishi
• 通讯作者: Atsuo Takanishi