Abstract: The forward displacement analysis problem of TT VGT manipulators with redundant degree of freedom is trained based on BP neural network, and then a solution to inverse displacement analysis problem is obtained. According to the above the network model, the first and the second order influence coefficients are derived, and the pseudo inverse of Jacobian matrix is obtained by using neural network. Finally the simulation calculation of kinematics for a seven celled tetrahedron tetrahedron variable geometry truss manipulator is given for illustration.