A new analytic model for the inverse kinematics of redundant robot is developed based on optimization measure that is defined as a criterion to determine the optimization adaptability. A smoothly continuous function of optimization measure is derived. The redundant seven-DOF robot is kinematically simulated using MATLAB software. Results show that self-motion of joints can vary continuously and the joint angular velocity can be controlled within the smaller range by using the new model for the inverse kinematics of redundant robot.