会员体验
专利管家(专利管理)
工作空间(专利管理)
风险监控(情报监控)
数据分析(专利分析)
侵权分析(诉讼无效)
联系我们
交流群
官方交流:
QQ群: 891211   
微信请扫码    >>>
现在联系顾问~
热词
    • 1. 发明专利
    • Control device, control system, and program
    • 控制装置,控制系统和程序
    • JP2008225607A
    • 2008-09-25
    • JP2007059672
    • 2007-03-09
    • Institute Of Physical & Chemical ResearchToyota Motor Corpトヨタ自動車株式会社独立行政法人理化学研究所
    • SHIMODA SHINGOKIMURA HIDENORIYAMAOKA MASAAKISUGIHARA HISAYOSHI
    • G06N3/00G06N3/04
    • PROBLEM TO BE SOLVED: To provide a control device for easily controlling various control objects by introducing a survival principle to a neuron simulating a nerve cell, and combining basic structures. SOLUTION: In a control device 501, an input layers 502, a learning layer 504 and an output layer 505 in which a plurality of neuron parts 101 based on a bursting model are arranged are connected by a weight α[i, j] and a weight β[j, i], and an update part 507 updates and learns the weight α[i, j] based on a Hebb rule from the corresponding relation of the ignition or an output of the input layer 503 and the output layer 505, and a control input such as a target value or an actual measurement value is applied through a control input part 502 to the input layer 503, and a control output part 508 outputs a quantity proportional to the number of pieces of the ignition of the output layer 505 as a control output applied to a control object, and initial values of variables corresponding to thresholds of ignition are made different in respective neuron parts 101 in the input layer 503, the learning layer 504, and the output layer 505. COPYRIGHT: (C)2008,JPO&INPIT
    • 要解决的问题:提供一种控制装置,用于通过向模拟神经细胞的神经元引入生存原理并组合基本结构来容易地控制各种控制对象。 解决方案:在控制装置501中,输入层502,学习层504和输出层505,其中布置有基于突发模型的多个神经元部分101通过权重α[i,j ]和权重β[j,i],并且更新部分507根据来自点火的对应关系或输入层503的输出和输出的对应关系,基于Hebb规则更新并学习权重α[i,j] 层505,并且诸如目标值或实际测量值的控制输入通过控制输入部分502施加到输入层503,并且控制输出部分508输出与点火次数成比例的量 作为施加到控制对象的控制输出的输出层505和对应于点火门限的变量的初始值在输入层503,学习层504和输出层505中的各个神经元部分101中不同。 P>版权所有(C)2008,JPO&INPIT
    • 3. 发明专利
    • Cooperative work robot and its control method
    • 合作工作机器人及其控制方法
    • JP2008213119A
    • 2008-09-18
    • JP2007056681
    • 2007-03-07
    • Institute Of Physical & Chemical ResearchToyota Motor Corpトヨタ自動車株式会社独立行政法人理化学研究所
    • TYTUS WOJTARASHIMODA SHINGOKIMURA HIDENORIUCHIHARA MASAFUMIMAEDA IWAOMURAYAMA HIDEYUKI
    • B25J13/08
    • PROBLEM TO BE SOLVED: To provide a cooperative work robot having an operator (a person) as a subject of an operation and efficiently carrying and exactly positioning an object difficult to be carried and positioned by only a person along an intention of the one operator by a cooperative work with the operator.
      SOLUTION: This cooperative work robot carries and positions the object 1 difficult to be carried and positioned by one person along the intention of the person by the cooperative work with the person. The cooperative work robot is provided with: a three-dimensionally movable robot hand 10 holding one end part of the object 1 via a free joint 12; a torque sensor 17 for detecting torsion torque acted by the person holding the other end part of the object 1; a thrust sensor 15 for detecting thrust force acing on the robot hand 10 from the other end part; an angle sensor 14 for detecting a displacement angle in relation to the robot hand 10 of the object 1; and a control device 20 for detecting the intention of the person from the torsion torque, the thrust force, and the displacement angle so as to control a three-dimensional position of the robot hand 10.
      COPYRIGHT: (C)2008,JPO&INPIT
    • 要解决的问题:提供一种具有作为操作对象的操作者(一个人)的合作作业机器人,并且仅有一个人能够高效地携带和精确地定位难以承载和定位的对象, 一个操作员通过与操作员的合作工作。 解决方案:该合作工作机器人通过与该人的合作工作,沿着人的意图携带并定位一个人难以携带和定位的对象1。 协同作业机器人具有:通过自由接头12保持物体1的一个端部的三维可移动的机械手10; 用于检测由保持物​​体1的另一端部的人作用的扭矩的扭矩传感器17; 用于检测来自另一端部的机械手10的推力的推力传感器15; 用于检测相对于物体1的机器人手10的位移角度的角度传感器14; 以及控制装置20,用于根据扭矩,推力和位移角检测人的意图,以便控制机器人手10的三维位置。(C)2008年, JPO&INPIT
    • 4. 发明专利
    • Neuronal apparatus, neuronal network apparatus, nonnegative integer encoder, integer cluster unit, feedback controller, and program
    • 神经网络设备,神经网络设备,非整数整数编码器,整数集群单元,反馈控制器和程序
    • JP2009087114A
    • 2009-04-23
    • JP2007257123
    • 2007-10-01
    • Institute Of Physical & Chemical ResearchToyota Motor Corpトヨタ自動車株式会社独立行政法人理化学研究所
    • SHIMODA SHINGOKIMURA HIDENORIYAMAOKA MASAAKIKAJIMA HIDEKI
    • G06N3/08G05B13/02G06N3/00
    • G06N3/063G05B2219/39031G06N3/04
    • PROBLEM TO BE SOLVED: To provide a neuronal apparatus and the like suitable for performing appropriate control by making a neuron threshold variable, extending Hebb's rule, and devising integer expressions, in a neuronal network.
      SOLUTION: In the neuronal apparatus 101 for simulating nerve cells, a threshold storage part 102 stores a threshold variable θ and two threshold coefficients Δθ1 and Δθ2, an input reception part 103 receives one or more input of input signal values at prescribed time intervals, an output part 104 outputs as an output signal value 1 assuming that the neural apparatus 101 is igniting when a total sum s of input signal values is threshold variable θ or larger and 0 assuming that the neural apparatus 101 is at rest otherwise; and a threshold update part 105 calculates Δθ1X + Δθ2(X-1) from an outputted output signal value X and the stored threshold coefficients Δθ1 and Δθ2 and updates the value of the threshold variable θ to be stored by the threshold storage part so that it increases just by the result of calculation.
      COPYRIGHT: (C)2009,JPO&INPIT
    • 要解决的问题:提供适合于通过在神经元网络中进行神经元阈值变量,延伸Hebb规则和设计整数表达来执行适当控制的神经元装置等。 解决方案:在用于模拟神经细胞的神经元装置101中,阈值存储部分102存储阈值变量θ和两个阈值系数Δθ1和Δθ2,输入接收部分103在规定时间接收输入信号值的一个或多个输入 假设神经装置101正在点燃输入信号值的总和为阈值变量θ或更大时,输出部分104输出为输出信号值1,假设神经装置101处于静止状态; 并且阈值更新部件105根据输出的输出信号值X和存储的阈值系数Δθ1和Δθ2计算Δθ1X+Δθ2(X-1),并且更新由阈值存储部分存储的阈值变量θ的值,使得 仅靠计算结果增加。 版权所有(C)2009,JPO&INPIT
    • 5. 发明专利
    • Power assist device and its control method
    • 电力辅助装置及其控制方法
    • JP2009039814A
    • 2009-02-26
    • JP2007207152
    • 2007-08-08
    • Institute Of Physical & Chemical ResearchToyota Motor Corpトヨタ自動車株式会社独立行政法人理化学研究所
    • ODAJIMA TADASHIUCHIHARA MASAFUMITYTUS WOJTARASHIMODA SHINGOKIMURA HIDENORI
    • B25J3/00B66F19/00
    • PROBLEM TO BE SOLVED: To provide a power assist device controlling the position and posture of a workpiece by providing a virtual guide as an area which the fingertip of a robot or the workpiece cannot enter, and by determining the end-point velocity of the robot.
      SOLUTION: This power assist device includes: a robot arm 3 having a plurality of revolute joints 3a; a suction tool 5 three-dimensionally swingably connected to the robot arm 3 through a free joint 4 and holding a window 2; an encoder 10 detecting the angles of the revolute joints 3a and free joint 4; a first calculation section 8a calculating the position of the fingertip 3c of the robot arm 3 or the position of the window 2 based on the angles; a storage section 8c storing the virtual guide G as the area not admitting the entering of the window 2; a second calculation section 8b calculating a velocity command value for putting restrictions not to exceed the virtual guide G by comparing the position of the fingertip 3c or the position of the window 2 with the position of the virtual guide G; and control means 8 controlling the fingertip 3c by the velocity command value.
      COPYRIGHT: (C)2009,JPO&INPIT
    • 要解决的问题:提供一种通过提供作为机器人或工件不能进入的区域的虚拟引导件来控制工件的位置和姿势的动力辅助装置,并且通过确定终点速度 的机器人。 解决方案:该动力辅助装置包括:具有多个旋转接头3a的机器人手臂3; 通过自由关节4三维可摆动地连接到机器人手臂3并保持窗口2的吸引工具5; 检测旋转接头3a和自由接头4的角度的编码器10; 第一计算部分8a,基于角度计算机器人手臂3的指尖3c的位置或窗口2的位置; 存储虚拟导轨G作为不允许进入窗口2的区域的存储部8c; 第二计算部分8b通过将指尖3c的位置或窗口2的位置与虚拟引导件G的位置进行比较来计算用于施加不超过虚拟引导件G的限制的速度指令值; 以及通过速度指令值控制指尖3c的控制装置8。 版权所有(C)2009,JPO&INPIT
    • 8. 发明专利
    • Work assisting device
    • 工作辅助设备
    • JP2008036777A
    • 2008-02-21
    • JP2006214618
    • 2006-08-07
    • Institute Of Physical & Chemical ResearchToyota Motor Corpトヨタ自動車株式会社独立行政法人理化学研究所
    • UCHIHARA MASAFUMIMAEDA IWAOYAMAOKA MASAAKIMURAYAMA HIDEYUKIKIMURA HIDENORITYTUS WOJTARANAKAYAMA SATOYUKISHIMODA SHINGO
    • B25J13/08
    • PROBLEM TO BE SOLVED: To provide a work assisting device that assists a worker to shift an object. SOLUTION: The work assisting device 100 is provided with: a gripping device 10; and a shifting device 20 that is connected to the gripping device 10 in a manner to be rotatable around a rotation axis Q. The object W is held by the gripping device 10 and a worker T. When the object W is rotated by the worker T, the work assisting device 100 calculates a first shifting command value that directs shifting direction and distance of the rotating axis Q for giving the rotation to the object W presumably to minimize deviation between the object W and a target attitude angle. The shifting device 20 is shifted according to the first shifting command value. The gripping device 10 and the shifting device 20 are revolvably connected, so that a rotation of the object W when shifting the rotating axis Q is defined by a positional relation of the rotating axis Q and the worker T. The worker T easily understands the rotation of the object W when the rotating axis Q is shifted. The worker T shifts the object W using the work shifting device 100 in easy understanding for the worker. COPYRIGHT: (C)2008,JPO&INPIT
    • 要解决的问题:提供一种帮助工人移动物体的工作辅助装置。 解决方案:工作辅助装置100设置有:夹持装置10; 以及以能够围绕旋转轴线Q旋转的方式与夹持装置10连接的换档装置20.被夹持装置10和作业者T保持物体W.当被加工物T被工人T旋转时 工作辅助装置100计算引导用于给予对象W的旋转轴线Q的移动方向和距离的第一移动指令值,以便最小化物体W与目标姿态角度之间的偏差。 移位装置20根据第一移位指令值而移位。 夹持装置10和移动装置20可旋转地连接,使得当旋转轴线Q移动时物体W的旋转由旋转轴线Q和工人T的位置关系限定。工人T容易理解旋转 当旋转轴线Q偏移时,物体W。 工人T使用工作移动装置100对工件的容易理解来移动对象W。 版权所有(C)2008,JPO&INPIT
    • 9. 发明专利
    • Power assist device and its adaptive model prediction control method
    • 电力辅助装置及其自适应模型预测控制方法
    • JP2007076807A
    • 2007-03-29
    • JP2005266438
    • 2005-09-14
    • Institute Of Physical & Chemical ResearchToyota Motor Corpトヨタ自動車株式会社独立行政法人理化学研究所
    • NAKAYAMA SATOYUKITYTUS WOJTARASHIMODA SHINGOKIMURA HIDENORIFUJIMOTO HIDEOUCHIHARA MASAFUMIMAEDA IWAOYAMAOKA MASAAKIMURAYAMA HIDEYUKI
    • B66F19/00
    • PROBLEM TO BE SOLVED: To provide a power assist device and its adaptive model prediction control method for suppressing deviation from a target final position to small extent as much as possible by smallest possible operation force in cooperation work of a man and a robot.
      SOLUTION: This power assist device is provided with an articulated arm robot 10 having a fingertip part 11 for gripping an object 1 to be transported rigidly and a controller 20 for controlling the robot. The robot 10 has a sensor for detecting revolute joint angle vector q of multi-revolute joint, a fingertip part position x, and force f
      e that a worker working in cooperation with the robot applies on the object to be transported. The controller 20 infers coefficient of viscosity of worker's fingertip D
      e , its modulus of elasticity K
      e , and a natural length position x
      e of worker's fingertip based on the revolute joint angle vector q, the fingertip part position x, and worker's force f
      e detected by the sensor on real time. The robot 10 generates assist force to suppress positioning error to small extent as much as possible by smallest possible operation force based on the inferred values.
      COPYRIGHT: (C)2007,JPO&INPIT
    • 要解决的问题:提供一种动力辅助装置及其自适应模型预测控制方法,用于通过人和机器人的协作工作中最小的可能的操作力尽可能多地抑制从目标最终位置到少量的偏移 。 解决方案:该动力辅助装置设置有铰接臂机器人10,其具有用于夹持要刚性运输的物体1的指尖部分11和用于控制机器人的控制器20。 机器人10具有用于检测与机器人协作工作的工作人员对物体进行多重旋转接头的旋转关节角度矢量q,指尖部位置x和力f e < 被运送 控制器20推测工人的指尖D e 的粘度系数,其弹性模量K e 和工人的自然长度位置x e 基于由传感器实时检测的旋转关节角度矢量q,指尖部位置x和作业者力f e 的指尖。 机器人10产生辅助力,以通过基于推断值的最小可能的操作力尽可能多地抑制定位误差。 版权所有(C)2007,JPO&INPIT