Automatioи of movement of the anthropomorphic systems is щи important task. It requires development of various motion control algorithms. Control of the exoskeleton is typically done using the reader and gain of the control pulses of a person. However, such a control cannot be implemented, for example, for the recovery of the motor abilities after injuries of the vertebrae and for learning to walk again, or in sports training with the help of the exoskeleton or its parts. In this case, the analytical motion control algorithms are required. The article presents a comparative analysis of these two approaches to the task of the motion control of the exoskeleton. As an illustration, the authors present a graphical dependence of the basic kinematic and dynamic parameters of the walking time and footage of the split-pictorial visualization of a model of an exoskeleton. The difference of this work from the existing ones is in the use of the exoskeleton links of a variable length. The paper describes a model of a link of a variable length in the form of a weightless rod and three lumped masses located at the ends of the rod, and an arbitrary point of the rod between its ends. The masses in the hinges-joints, except the movable joints, can be modeled by a motor mounted on a hinge pin having a substantial mass and creating the control point. Giving each lot a certain value, it is possible to obtain such a distribution of the masses on the rod, which approximately corresponds to the inertial properties of the link in the musculoskeletal system of a person. The differential equations of motion for the single-link and multi-chain systems were obtained. Generalizations for development of an effective matrix and recursive algorithms for writing differential equations of motion for the multi-link systems composed of links of variable length were composed. A study revealed that control of a real human walk had a pronounced pulsed character, while a gait, based on the theoretically specified periodic functions, was more energy intensive than a real human walk. This, in particular, explains why the energy consumption of the modern anthropomorphic robots during walking is higher than that of humans.