Algorithms of Self-control of the Underwater Walking Device According to Information on a Collision of Feet with an Unorganized Support Surface

In underwater conditions walking machines have better ground and shape passability. Control of almost all robotic systems moving on a bottom (underwater bulldozers, cable layers, self-moved ground mining units, etc.) is executed from a surface by the operator according to the visual information from onboard video sensors. Such approach to control of the multi-legged walking machines is not appropriate. The operator cannot to control effectively all legs at the same time and does not manage to make good decisions at emergence of any obstacle in the direction of movement, because of poor visibility. Results of investigation of information opportunities of foot of the cyclic walking mover during collision with an unorganized support surface are discussed. Shown that on the type of self-adapting of foot to an unorganized support surface during collision with an obstacle and the relative movement of the mechanism of walking it is possible to obtain some indistinct information about external working space and the current situation. It can be used at adoption of the correct decisions on control in the conditions of absence of external control signal. In particular, on the basis of fuzzy control algorithms, realization of the standard typical movements which are independently executed by the underwater walking robot for an exception of capsizing and other emergencies is possible.