欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
提出了用人工神经元网络（ANN）补偿大射电望远镜（LT）中柔索驱动并联机器人（WDPR）系统动平台位姿误差的方法。为了提高WDPR的运动精度，建立了3种可行的误差补偿方案，并运用kevenberg．Marquart（L-M）算法训练了相应的3个神经元网络。标定仿真显示，基于索长补偿的柔性标定方案比基于动平台位姿补偿的标定方案好。研究结果为提高LT舱索系统的控制精度奠定了理论基础。
英文摘要：
A compensating method for the position and posture errors of the moving platform of a wire driven parallel robot （WDPR） system for the large radio telescope （LT） was presented by artificial neural networks （ANN）. Three possible schemes were modeled to improve the moving accuracy of WDPR, and the corresponding ANN nets were trained by the Levenberg-Marquart algorithm. The calibration simulations showed that the flexible calibration scheme based on cablelength compensation has more advantages over the scheme based on moving platform-position compensation. The results lay the theoretical foundation for improving the control accuracy of the cable-cabin system for the large radio telescope.
同期刊论文项目
同项目期刊论文
期刊信息
位置：