Title:
LEARNING CONTROLLER
Document Type and Number:
Japanese Patent JP3191836
Kind Code:
B2
Abstract:
PURPOSE: To attain the follow-up control with less operation quantity by receiving an object command at a current time and an output of a controlled system before sampling so as to allow the output of the controlled system to follow up an object command repeating a same pattern thereby utilizing directly a transfer function model.
CONSTITUTION: The controller receives a current value r(i) of an object command repeating a same pattern for a period L and an output y(i-D) of a controlled system before D-sampling and provides an output of a control input u(i) to the controlled system. A subtractor 8 obtains a difference e(i-D) between an object command r(i-D) stored in a memory 2 and the output y(i-D). A subtractor 7 calculates a correction rate σ(i) and it is given to an adder 9 and stored in a memory 5. The adder 9 adds the output σ(i) of the adder 7 and the value u (i') stored in a memory 6 to calculate the control input u(i). The obtained control input u(i) is outputted to the controlled system as an output of the learning controller 1 and stored in the memory 6.
Inventors:
Yuji Nakamura
Application Number:
JP13546993A
Publication Date:
July 23, 2001
Filing Date:
May 12, 1993
Export Citation:
Assignee:
Yaskawa Electric Co., Ltd.
International Classes:
G05B13/02; G05B13/04; G05B21/02; (IPC1-7): G05B13/02; G05B13/04; G05B21/02
Domestic Patent References:
| JP1237701A | ||||
| JP4372001A | ||||
| JP4358202A |