http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
오프라인 로봇 작업 프로그래밍을 위한 Nuero-Net 기반의 직접 교정
국금환,황헌 慶尙大學校 工科大學 自動化및컴퓨터應用技術硏究所 1997 自動化 및 컴퓨터應用技術 Vol.4 No.1
First, robot task error was analyzed. Considering the practical implementation of robot task calibration, the hybrid type of calibration is proposed. Kinematic errors of robot model parameters and geometric errors of workpart position ore analytically identified using forward calibration based on the indirect measurement. The neuro-net based inverse calibration is performed for compensating dynamic effect of robot motion and for generating directly compensated joint moves. PUMA 560 was simulated for compensation verification. For robot tasks cooperated with computer vision, the neuro-net based scheme was also applied to compensate the geometrically distorted camera input image. The compensation effect was verified using the known sized geometric primitives. Cerebellar Model Linear Associator Net(CMLAN), a kind of an adaptive control function generator, was successfully utilized for a calibration.