Neural network approach to condition monitoring of industrial robotic-manipulators
Мogilev State University of Food Technologies
The paper presents a novel technique for condition monitoring of industrial robotic manipulators, which is based on neural network analyses of the dynamic model parameters obtained by means of on-line identification. There were derived analytical expressions that allow minimising impact of the measurement
errors on the identification accuracy. Efficiency of the proposed technique has been verified by real-life case studies from industrial monitoring systems.
condition monitoring, robotic-manipulators, neural network
For citing this article
Kazheunikau M.M. Neural network approach to condition monitoring of industrial robotic-manipulators // The Research of the Science City, 2013, no. 1, pp. 42-48.