Control algorithms for motors of anthropomorphic gripping with group drive

Problem statement: the current trend is the use in various spheres of human activity of service robots equipped with anthropomorphic grippers (AG) with the capabilities inherent in the human hand. The functionality of the AG is determined by the executive groups of links (IGZ) having a group drive with a variable structure. When enclosing an external object (VO), each opposed IGZ sequentially implements a contact with three output links. The algorithm for the simultaneous control of the motors must ensure that the position of the unsecured AO remains unchanged. Known AG control methods, for example impedance, do not provide this requirement. In addition, possible methods are focused on an unchanging control object, which is one output link of the IGZ. The variability of control objects, characteristic of systems with a group drive and a variable structure, determines the need for a fundamentally new approach to the control of engines of opposed IGZs. Methods used: theoretical research is based on the main provisions of the analysis of the functioning of complex systems, information processing that determines their state and decision-making based on them. The novelty of the proposed algorithms lies in the fact that the engine control is based on information about the interaction of the output links of both this IGZ and the opposite one. The sequence of switching on the motors is determined based on the analysis of changes in the set of conditions. External – the contact of the current control object with the VO, internal – an increase in the torque on the engine in comparison with the value that determines the free movement. This approach makes it possible, with only contact sensors on the output links, to implement adaptive control of the motors of the opposed IGZ. Result: The use of the proposed algorithms makes it possible to formalize the control of the motors of the opposed IGZ taking into account the position of the initially non-deterministic AO. Practical significance: the developed algorithms are designed to control anthropomorphic grips performing actions with the AO in unfavorable conditions for humans and the uncertainty of the AO position. Their use will increase the functionality of robots equipped with AG.