Path planning and redundancy resolution for planar redundant manipulators, using artificial neural networks

Given an end-effector position there are infinite corresponding joint configurations for redundant manipulators, which can be represented in terms of a finite set of self-motion manifolds in the configuration space. This paper presents an artificial neural network based software tool to solve redundancy resolution and path planning problems simultaneously, using self-motion topology knowledge of redundant manipulators. The neural network architecture auto-configures its structure according to the problem it is solving such that the number of neurons and the number of inputs for each neuron vary between different problems. The architecture is also suitable for distributed computing or multiprocessing implementations