Analysis of Walking Stability of Robot using Neural Network algorithms as an Activation Function of the System Characteristic Curve

Abstract

Background/Objectives: The robots is to create machines that collectively mimic human and physical behaviors. However, robots are complexed in the real-time walking control of robot. If such a complex expression was included in control operation, it leads to the disadvantage that operation time has been lengthened.

Methods/Statistical analysis: It was applied neuron networks algorithm for the real-time walking control of Multi-articulated robot. It could be a new method to implement a neural network controller by installing a real object for controlling and an algorithm. A new method was study to the existing method of implementing a neural network controller by utilizing activation function at the output node.

Findings: Multi-articulated robot is expressed with a complicated mathematical model on account of the mechanic, electric non-linearity which each articulation of mechanism has, and includes an unstable factor in time of walking control. So, if the rapid change of the load or the disturbance is given, it is difficult to fulfill the control of desired performance.

Improvements/Applications: The proposed control algorithm generated control signs corresponding to the non-linearity of Multi-articulated robot, which could generate desired motion in real time. 

Keywords: Humanoid, Multi-articulated Robot, neural Network, Real-Time, Walking, Mechanic