Response Surface Methods Study for Industrial Robot Natural Frequency Prediction

Natural frequency information of industrial robot can be used for vibration suppression in robot configuration optimization. Robot natural frequency in whole working space is able to be predicted by a mathematical model fitted with the data of several sample configurations selected through Response Surface Methods (RSM). As there exists different types of RSMs, this paper aims at choosing an appropriate method for sample selection. Through analysis and practical experiments, Central Composite Inscribed Design is indicated to be the optimal one.