Hyper-Chaotic Mathematical Programming Method and Its Application to Dodecahedron Variable Geometry Truss Manipulator

As a new robot mechanism variable geometry truss manipulator can be applied widely for its good performance. Forward displacement of variable geometry truss manipulator is always transformed into finding the solutions of nonlinear equations. Newton iterative method is an important technique to one dimensional and multidimensional variable and often used to solve nonlinear equations. Iterative process is sensitive to the initial point. The mathematical programming method is adopted when the iteration diverges with the Newton and quasi-Newton methods. A new method of finding all solutions of nonlinear questions is proposed, in which we combine mathematical programming method with hyper-chaotic neuron network system and utilize hyper-chaotic network system to obtain locate initial points. The numerical example in dodecahedron variable geometry truss manipulator synthesis shows that all solutions have been quickly obtained, and it also shows that the method is correct and effective. This provides a simple and new method for mechanism design.