Solving model of multi-degree of freedom of prosthesis joint space

Abstract

In this study, an implementation of solution for solving multi-degree of freedom prosthesis joint space is introduced, which is in the case of only knowing three location parameters information of random target in the operating space. Utilizing unique parameters mapping capability for incomplete information, fast parallel information processing capacity and multiple-input multiple-output nonlinear characteristics of artificial neural network establishes a solving model of multi-degree of freedom prosthesis joint space parameters. When the prosthesis system obtains the objectives' three-dimensional coordinates value in the operating space, taking the three coordinate values and a less defined compensatory attitude description as the input of network model, you can obtain a feasible solution of all the prostheses joint quickly and in parallel. This method has been verified with the instance and simulation.