EP0377467A2
Expired / Lapsed
EP
⏱️ 0.0 years remaining
Movement trajectory generating method of a dynamical system
Assignee: ATR AUDITORY AND VISUAL PERCEPTION RESEARCH LABORATORIES
•
Filing: 1990-01-05
•
Granted: —
•
Expires: 2010-01-05
Abstract
Abstract A movement trajectory generating system of a dynamical system uses neural network units (1, 2, 3) including cascade connection of a first layer (11, 21, 31), a second layer (12, 22 32), a third layer (13, 23, 33) and a fourth layer (14, 24, 34), to learn vector field of differential equations indicating forward dynamics of a controlled object (4). Conditions concerning trajectories of a final point and a via-point of movement of the controlled object and locations of obstacles are given from a motor center (5). While smoothness of movement is ensured by couplings of electric synapses using errors with respect to those conditions as total energy, least dissipation of energy is attained, whereby trajectory formation and control input for realizing the trajectory are obtained simultaneously.
Inventors
- 👤 Mitsuo Atr Auditory And Visual Perception Kawato
- 👤 Yoshiharu Int. Inst. Advanced Study Social Maeda
- 👤 Yoji C/O Faculty Of Engineering Of Uno
- 👤 Ryoji C/O Faculty Of Engineering Of Suzuki
Bibliographic Details
Application Number
—
Jurisdiction
EP
Data Source
google_patents
Last Synchronized
2026-11-06 13:42
CPC Classifications
G; G05; G05B; G05B19/00; G05B19/02; G05B19/18; G05B19/41; B; B25; B25J; B25J9/00; B25J9/16; B25J9/1602; B25J9/161; G06; G06N; G06N3/00; G06N3/02; G06N3/04; G06N3/045; G06N3/0499; G06N3/08; G06N3/09; G05B2219/00; G05B2219/30; G05B2219/33; G05B2219/33027; G05B2219/42; G05B2219/42155; G05B2219/49; G05B2219/49157
Claims Analysis
Total Claims: 6
Claim 1
Independent Claim
120 words
A movement trajectory generating method of a dynamical system, for generating a time trajectory of a controlled object, said time trajectory of said controlled object being generated by using conditions for estimating the movement trajectory such as a movement time, a start point, an end point and a via-point of the trajectory, and obstacles to the movement, and using a criterion for ensuring smoothness of the movement trajectory by minimizing evaluation functions in integration for a movement time obtained by square of a jerk such as the third derivative of position with respect to time, integration for a movement time obtained by squaring time change of torque, and integration for a movement time obtained by squaring time change of force.
Claim 2
Dependent Claim
120 words
A movement trajectory generating method of a dynamical system, for generating a time trajectory of a controlled object, said time trajectory of said controlled object being generated by using conditions for estimating the movement trajectory such as a movement time, a start point, an end point and a via-point of the trajectory, and obstacles to the movement, and using a criterion for ensuring smoothness of the movement trajectory by minimizing evaluation functions in integration for a movement time obtained by square of a jerk such as the third derivative of position with respect to time, integration for a movement time obtained by squaring time change of torque, and integration for a movement time obtained by squaring time change of force.
Claim 3
Dependent Claim
80 words
A movement trajectory generating method of a dynamical system in accordance with claim 1, wherein a number of internal models obtained by multiplying by a unit of time a vector field of nonlinear differential equations describing dynamical characteristics of said controlled object are cascade-connected, and the time trajectory of said controlled object is generated by using bypasses corresponding to a flow of the differential equations and discrete diffusion couplings for satisfying the evaluation criterion for ensuring the smoothness of movement.
Claim 4
Dependent Claim
87 words
A movement trajectory generating method of a dynamical system in accordance with claim 2, wherein a multilayer neural network for learning the internal models obtained by multiplying, by the unit of time, the vector field of said nonlinear differential equations is included, a large number of said multilayer neural networks being cascade-connected, and said time trajectory of said controlled object is generated by using bypasses corresponding to a flow of the differential equations and electric synapses for satisfying the evaluation criterion for ensuring the smoothness of movement.
Claim 5
Dependent Claim
104 words
A movement trajectory generating method of a dynamical system, for generating a time trajectory of a controlled object, a number of cascade-connected multilayer neural networks being included, said method comprising learning forward dynamics and kinematics of said controlled object by using said multilayer neural networks, and providing conditions concerning trajectory of a final point and a via-point of movement of said controlled object, and locations of obstacles, generating the trajectory and obtaining a control input for realizing the trajectory simultaneously by least dissipation of energy while ensuring smoothness of movement by electric synaptic couplings using errors with respect to said conditions as total energy.
Claim 6
Dependent Claim
104 words
A movement trajectory generating method of a dynamical system, for generating a time trajectory of a controlled object, a number of cascade-connected multilayer neural networks being included, said method comprising learning forward dynamics and kinematics of said controlled object by using said multilayer neural networks, and providing conditions concerning trajectory of a final point and a via-point of movement of said controlled object, and locations of obstacles, generating the trajectory and obtaining a control input for realizing the trajectory simultaneously by least dissipation of energy while ensuring smoothness of movement by electric synaptic couplings using errors with respect to said conditions as total energy.
References Cited (Backward Citations)
No citation records indexed.
USPTO Annuity & Maintenance Windows
Maintenance schedule is calculated for granted US utility patents.