Templates of ship movement anthropomorphic control: design and application

Abstract

The process of designing templates of anthropomorphic ship movement control is described in the paper. The solutions of the sequence of modeling problems are used in the state space of the ergatic "skipper-ship" system for the example of a ship of "Volga-Don" type. Parametric interval uncertainties are taken into account in the initial mathematical model of the controlled object. They influence the unambiguity of solving the optimal control problem in the classical formulation. Factors of a priori uncertainty are the draft of the ship and the depth of the ship's passage. Numerical solutions to the problem of optimal ship movement control in terms of speed are obtained for intervals ends of the mathematical model parameters values. The solutions are used in the design of templates to describe and account for the uncertainty of the allocation of management resources according to the methodology of interaction planning in an ergatic system. The templates of anthropomorphic control in the skipper-ship system are represented by sequences of element numbers of a set of normal systems of ordinary differential equations. They are constructed in a five-dimensional space of states according to a transformed mathematical model of the ship. Each normal system of ordinary differential equations displays in mathematical form an incomplete representation of the action of virtual controls and the corresponding elementary motion of the ship due to the existence of uncertainties. The procedures for constructing various options for the allocation of anthropomorphic control resources and corresponding templates of anthropomorphic control according to expert estimates based on a variety of solutions to the optimal control problem are proposed. The constructive property of the templates of anthropomorphic control is illustrated by the specific examples. New templates can be built by joining standardtemplates in a certain sequence, and the standard templates themselves can be determined using information about the performed movements of the ship controls, taking into account the experience of navigation. It is shown that it is possible to use such a posteriori information to train the automatic control machine of the "skipper-ship" system by rational management methods, including those that cannot be obtained as solutions to mathematical problems of optimal control. The library of anthropomorphic control templates is presented as an integral part of the knowledge base when using the technology of expert systems in the construction of automatic control machine with artificial intelligence.