A new trajectory planning method of manipulators with time scaling neural network

In this paper, a new trajectory planning method with a nonlinear optimization technique is proposed to search an optimal space path and time trajectory simultaneously. However, the proposed trajectory planning method for real industrial manipulators is not effective because the computational burden to calculate the inverse dynamics of manipulator increases exponentially as the degree of freedom increases. To reduce the computational burden, we introduced two neural networks that learn the inverse dynamics of manipulators and the time scaling which is the expansion and contraction of the time axis. As a result of using these neural networks, it has become possible to apply the planning method to plan an optimal trajectory in a shorter time than using a mathematical model.