机构地区: 韶关学院物理与机电工程学院
出 处: 《机床与液压》 2009年第12期47-51,共5页
摘 要: 研究了支持向量机回归算法,并将该算法运用于机器人绝对误差的补偿。针对机器人绝对误差补偿多输入、多输出的特点,提出了用多个支持向量机回归算法实现多输出的解决策略;提出了基于支持向量机回归算法误差的2种补偿方法,并进行了仿真与实验,验证了所提方法的有效性。仿真结果表明运用该方法可以使机器人的定位精度平均提高95%以上。 The support vector machine regression algorithm was studied, and it was used successfully in robot absolute error compensation. A multiple support vector solution strategy was proposed, and the problem of multiple inputs and multiple outputs of robot absolute error compensation was solved. Two error compensation methods were proposed based on support vector machine regression algorithm, and the validation was proved by simulation and tests. The simulation tests show that robot absolute error is improved over 95% in average.
关 键 词: 机器人 绝对误差 误差补偿 支持向量机 回归算法
领 域: [自动化与计算机技术] [自动化与计算机技术]