公开(公告)号：US08886357B2

公开(公告)日：2014-11-11

申请号：US13432094

申请日：2012-03-28

申请人： Norikazu Sugimoto ,  Yugo Ueda ,  Tadaaki Hasegawa ,  Soshi Iba ,  Koji Akatsuka

发明人： Norikazu Sugimoto ,  Yugo Ueda ,  Tadaaki Hasegawa ,  Soshi Iba ,  Koji Akatsuka

IPC分类号： G06F19/00 ,  G05B13/02

CPC分类号： G05B13/0265 ,  Y10S901/03

摘要： It is possible to perform robot motor learning in a quick and stable manner using a reinforcement learning apparatus including: a first-type environment parameter obtaining unit that obtains a value of one or more first-type environment parameters; a control parameter value calculation unit that calculates a value of one or more control parameters maximizing a reward by using the value of the one or more first-type environment parameters; a control parameter value output unit that outputs the value of the one or more control parameters to the control object; a second-type environment parameter obtaining unit that obtains a value of one or more second-type environment parameters; a virtual external force calculation unit that calculates the virtual external force by using the value of the one or more second-type environment parameters; and a virtual external force output unit that outputs the virtual external force to the control object.

摘要翻译： 可以使用强化学习装置以快速且稳定的方式执行机器人电动机学习，包括：获得一个或多个第一类型环境参数的值的第一类型环境参数获取单元; 控制参数值计算单元，其通过使用所述一个或多个第一类型环境参数的值来计算使奖励最大化的一个或多个控制参数的值; 控制参数值输出单元，其将所述一个或多个控制参数的值输出到所述控制对象; 获取一个或多个第二类型环境参数的值的第二类型环境参数获取单元; 虚拟外力计算单元，其通过使用所述一个或多个第二类型环境参数的值来计算所述虚拟外力; 以及将虚拟外力输出到控制对象的虚拟外力输出单元。

公开(公告)号：US08768507B2

公开(公告)日：2014-07-01

申请号：US13237018

申请日：2011-09-20

申请人： Soshi Iba ,  Tadaaki Hasegawa

发明人： Soshi Iba ,  Tadaaki Hasegawa

IPC分类号： G05B19/04

CPC分类号： B62D57/032

摘要： A robot and a behavior control system for the same are capable of ensuring continued stability while carrying out a specified task by a motion of a body of the robot. Time-series changing patterns of first state variables indicating a motional state of an arm are generated according to a stochastic transition model such that at least one of the first state variables follows a first specified motion trajectory for causing the robot to carry out a specified task. Similarly, time-series changing patterns of second state variables indicating a motional state of the body are generated according to the stochastic transition model such that the second state variables satisfy a continuously stable dynamic condition.

摘要翻译： 机器人及其行为控制系统能够通过机器人的身体的运动来确保持续的稳定性，同时执行指定的任务。 根据随机过渡模型产生指示臂的运动状态的第一状态变量的时间序列变化模式，使得至少一个第一状态变量遵循第一指定的运动轨迹，以使机器人执行指定的任务 。 类似地，根据随机过渡模型产生指示身体的运动状态的第二状态变量的时间序列变化模式，使得第二状态变量满足连续稳定的动态条件。

公开(公告)号：US08315740B2

公开(公告)日：2012-11-20

申请号：US12137873

申请日：2008-06-12

申请人： Tadaaki Hasegawa ,  Yugo Ueda ,  Soshi Iba ,  Darrin Bentivegna

发明人： Tadaaki Hasegawa ,  Yugo Ueda ,  Soshi Iba ,  Darrin Bentivegna

IPC分类号： G05B19/00

CPC分类号： G06N3/008 ,  G05B2219/40391

摘要： The present invention provides a motion control system to control a motion of a second motion body, by considering an environment which a human contacts and a motion mode appropriate to the environment, and an environment which a robot actually contacts. The motion mode is learned based on an idea that it is sufficient to learn only a feature part of the motion mode of the human without a necessity to learn the others. Moreover, based on an idea that it is sufficient to reproduce only the feature part of the motion mode of the human without a necessity to reproduce the others, the motion mode of the robot is controlled by using the model obtained from the learning result. Thereby, the motion mode of the robot is controlled by using the motion mode of the human as a prototype without restricting the motion mode thereof more than necessary.

摘要翻译： 本发明提供了一种运动控制系统，其通过考虑人类接触的环境和适合于环境的运动模式以及机器人实际接触的环境来控制第二运动体的运动。 基于这样的想法来学习运动模式，即仅仅学习人的运动模式的特征部分是足够的，而不需要学习其他动作模式。 此外，基于仅仅再现人的运动模式的特征部分而不需要再现其他的想法就足够了，通过使用从学习结果获得的模型来控制机器人的运动模式。 因此，通过使用人的运动模式作为原型来控制机器人的运动模式，而不限制其运动模式。

公开(公告)号：US20120078416A1

公开(公告)日：2012-03-29

申请号：US13237018

申请日：2011-09-20

申请人： Soshi Iba ,  Tadaaki Hasegawa

发明人： Soshi Iba ,  Tadaaki Hasegawa

IPC分类号： G05B19/04

CPC分类号： B62D57/032

摘要： A robot and a behavior control system for the same are capable of ensuring continued stability while carrying out a specified task by a motion of a body of the robot. Time-series changing patterns of first state variables indicating a motional state of an arm are generated according to a stochastic transition model such that at least one of the first state variables follows a first specified motion trajectory for causing the robot to carry out a specified task. Similarly, time-series changing patterns of second state variables indicating a motional state of the body are generated according to the stochastic transition model such that the second state variables satisfy a continuously stable dynamic condition.

摘要翻译： 机器人及其行为控制系统能够通过机器人的身体的运动来确保持续的稳定性，同时执行指定的任务。 根据随机过渡模型产生指示臂的运动状态的第一状态变量的时间序列变化模式，使得至少一个第一状态变量遵循第一指定的运动轨迹，以使机器人执行指定的任务 。 类似地，根据随机过渡模型产生指示身体的运动状态的第二状态变量的时间序列变化模式，使得第二状态变量满足连续稳定的动态条件。

公开(公告)号：US08014901B2

公开(公告)日：2011-09-06

申请号：US11989629

申请日：2006-07-10

申请人： Kuniaki Matsushima ,  Tadaaki Hasegawa ,  Yuichiro Kawaguchi

发明人： Kuniaki Matsushima ,  Tadaaki Hasegawa ,  Yuichiro Kawaguchi

IPC分类号： G05B15/00 ,  G05B19/00

CPC分类号： B62D57/032 ,  G05B2219/40195

摘要： In a mobile robot control system, it is configured such that the robot generates time-series data sequentially at a predetermined time interval and transmits them to the external terminal, and the external terminal receives the transmitted time-series data and adds them to the motion command, such that the motion of the robot is determined based on the generated time-series data and the time-series data added to the motion command. With this, it becomes possible to prevent the robot from suddenly starting to move at the time when the communication between the external terminal which is a transmitting source of the motion command and the robot has recovered from disconnection, thereby enabling to avoid making the operator feel unnatural.

摘要翻译： 在移动机器人控制系统中，其被配置为使得机器人以预定时间间隔顺序地生成时间序列数据并将其发送到外部终端，并且外部终端接收所发送的时间序列数据并将其相加到运动 使得基于生成的时间序列数据和添加到运动命令的时间序列数据来确定机器人的运动。 由此，能够防止机器人在作为运动指令的发送源的外部端子与机器人之间的通信从断开恢复时突然开始移动，从而能够避免使操作者感觉到 不自然

公开(公告)号：US20100114807A1

公开(公告)日：2010-05-06

申请号：US12610709

申请日：2009-11-02

申请人： Yugo Ueda ,  Tadaaki Hasegawa ,  Soshi Iba ,  Koji Akatsuka ,  Norikazu Sugimoto

发明人： Yugo Ueda ,  Tadaaki Hasegawa ,  Soshi Iba ,  Koji Akatsuka ,  Norikazu Sugimoto

IPC分类号： G06F15/18

CPC分类号： G06N99/005

摘要： A reinforcement learning system (1) of the present invention utilizes a value of a first value gradient function (dV1/dt) in the learning performed by a second learning device (122), namely in evaluating a second reward (r2(t)). The first value gradient function (dV1/dt) is a temporal differential of a first value function (V1) which is defined according to a first reward (r1(t)) obtained from an environment and is served as a learning result given by a first learning device (121). An action policy which should be taken by a robot (R) to execute a task is determined based on the second reward (r2(t)).

摘要翻译： 本发明的强化学习系统（1）利用由第二学习装置（122）执行的学习中的第一值梯度函数（dV1 / dt）的值，即评估第二奖励（r2（t）） 。 第一值梯度函数（dV1 / dt）是根据从环境获得的第一奖励（r1（t））定义的第一值函数（V1）的时间微分，并且被用作由 第一学习装置（121）。 基于第二奖励（r2（t））确定由机器人（R）执行任务应采取的动作策略。

公开(公告)号：US20090312867A1

公开(公告)日：2009-12-17

申请号：US12096987

申请日：2006-10-31

申请人： Tadaaki Hasegawa ,  Nobuyuki Ohno

发明人： Tadaaki Hasegawa ,  Nobuyuki Ohno

IPC分类号： G06F19/00

CPC分类号： B62D57/032

摘要： A provisional desired motion trajectory of an object is determined based on a moving plan of the object. Then, it is determined whether a robot leg motion can satisfy a necessary requirement. The requirement is related to a position/posture relationship between the object and the robot, and a determination of whether the requirement can be satisfied is made at a future, predetermined step. A restrictive condition related to robot leg motion is satisfied at each step up to the predetermined number of steps. If the requirement is satisfied, then a desired gait is generated on the basis of the provisional desired motion trajectory. Otherwise, a desired gait is generated on the basis of a desired motion trajectory of the object according to a corrected moving plan.

摘要翻译： 基于对象的移动计划来确定对象的临时期望运动轨迹。 然后，确定机器人腿部运动是否能够满足必要的要求。 该要求涉及物体与机器人之间的位置/姿势关系，并且在将来的预定步骤中确定是否满足要求。 在每一步达到预定步数的步骤时，满足与机器人腿运动有关的限制条件。 如果满足要求，则基于临时期望运动轨迹产生期望的步态。 否则，根据校正的移动计划，基于物体的期望的运动轨迹产生期望的步态。

公开(公告)号：US06505096B2

公开(公告)日：2003-01-07

申请号：US09319266

申请日：1999-08-16

申请人： Toru Takenaka ,  Tadaaki Hasegawa ,  Takashi Matsumoto

发明人： Toru Takenaka ,  Tadaaki Hasegawa ,  Takashi Matsumoto

IPC分类号： G06F1900

CPC分类号： B62D57/02

摘要： In a legged mobile robot, in particular a biped robot having arms, the control is conducted such that the dynamic balance is preserved so as to keep a stable posture, even when the robot is subject to unexpected reaction force from an object. It is configured such that the difference or error (i.e., moment about the central point of total floor reaction force) between the desired object reaction force, and the actual value is determined and is distributed to the desired body position/posture and the desired feet position/posture, and based thereon, robot link joints are controlled to be driven.

摘要翻译： 在有腿的移动机器人中，特别是具有臂的双足机器人，即使当机器人受到来自物体的意想不到的反作用力时，进行控制，使得保持动态平衡以保持稳定的姿势。 它被构造成使得期望的物体反作用力和实际值之间的差异或误差（即，总地板反作用力的中心点的力矩）被确定并且被分配到期望的身体位置/姿势和期望的脚 位置/姿势，并且基于此，机器人连杆关节被控制以被驱动。

公开(公告)号：US06401846B1

公开(公告)日：2002-06-11

申请号：US09630743

申请日：2000-08-02

申请人： Toru Takenaka ,  Takayuki Kawai ,  Hiroshi Gomi ,  Tadaaki Hasegawa ,  Takashi Matsumoto

发明人： Toru Takenaka ,  Takayuki Kawai ,  Hiroshi Gomi ,  Tadaaki Hasegawa ,  Takashi Matsumoto

IPC分类号： B62D5702

CPC分类号： B62D57/032 ,  B25J19/0091

摘要： In a biped walking robot having a body and two articulated legs each connected to the body through a hip joint and having a knee joint and an ankle joint, connected by a shank link, a knee pad is mounted on the shank link as a landing/shock absorbing means at a position adjacent to the knee joint which is brought into contact with the floor when coming into knee-first contact with the floor such that the knee joint is to be positioned at a location forward of the center of gravity of the robot in a direction of robot advance, while absorbing impact occurring from the contact with the floor. With this, the robot can be easily stood up from an attitude with its knee joint regions in contact with the floor. Moreover, when coming into knee-first contact with the floor, it can absorb the impact of the contact to protect the knee joint regions and the floor from damage.

摘要翻译： 在具有主体和两个铰接腿的双足步行机器人中，每个腿部通过髋关节连接到身体，并且具有通过柄连杆连接的膝关节和踝关节，膝关节垫作为着陆/ 冲击吸收装置位于与膝关节相邻的位置处，当与地板进行膝盖 - 第一次接触时，其与地板接触，使得膝关节位于机器人重心的前方位置 在机器人前进的方向上，同时吸收与地板接触发生的冲击。 这样，机器人可以容易地从与膝盖接合区域与地板接触的姿态站起来。 此外，当与地板进行膝盖首次接触时，它可以吸收接触的冲击，以保护膝关节区域和地板免受损坏。

公开(公告)号：US5594644A

公开(公告)日：1997-01-14

申请号：US65242

申请日：1993-05-20

申请人： Tadaaki Hasegawa ,  Toru Takenaka

发明人： Tadaaki Hasegawa ,  Toru Takenaka

IPC分类号： B25J5/00 ,  B25J9/10 ,  B62D57/032 ,  G05D1/02 ,  G06F7/70 ,  G05B19/04

CPC分类号： G05D1/0212 ,  B62D57/032 ,  G05D1/0272 ,  G05D2201/0217

摘要： A trajectory generation for a member such as a foot of a legged mobile robot. First, basic trajectories defining some typical motions of the foot including a constraint condition are established on a virtual plane or surface fixed on a coordinate system. The virtual plane is kept fixed on the ground until a time the foot is to be lifted. Then at this period free from the constraint condition, the coordinate system is displaced such that the virtual surface coincides with another point of the ground on which the foot is to be landed. A trajectory for a footrise to footfall is thus generated by combining the basic trajectories in the coordinate system and the amount of displacement of the coordinate system. Thus, the boundary conditions become extremely simple and hence, trajectory generation is greatly simplified. A real time trajectory correction can be conducted if desired.

摘要翻译： 用于诸如腿式移动机器人的脚的构件的轨迹生成。 首先，在固定在坐标系上的虚拟平面或表面上建立了定义包括约束条件的脚的典型运动的基本轨迹。 虚拟平面保持固定在地面上，直到脚被抬起。 然后在这个时间没有约束条件的情况下，坐标系被移位，使得虚拟表面与脚将要降落的地面的另一个点重合。 因此，通过组合坐标系中的基本轨迹和坐标系的位移量来生成脚踏脚步的轨迹。 因此，边界条件变得非常简单，因此大大简化了轨迹生成。 如果需要，可以进行实时轨迹校正。