Instruments and Systems: Monitoring, Control, and Diagnostics Annotation << Back The application of the steepest descent method in a single bioinspired algorithm I.V. EVDOKIMOV, YE. D. KULAKOV At present, the question remains relevant research and development of the limits of applicability of certain groups of algorithms that are united by the fact that the mechanisms underlying their activities were derived from the study of processes occurring in the “living” world. Such algorithms are called bioinspired algorithms. Use variety of gradient methods is the method of steepest descent-complements the well-known algorithm description of the Shepherd’s dog. Work Shepherd with grazing sheep - a typical example, when one agent leads many to move in a given direction. This approach is widely applicable in practice to control the crowd, environmental cleanup Wednesday and other engineering tasks. Despite the fact that heuristic algorithm of the Shepherd’s dog has already been described in the scientific literature, the question remains, what is the algorithm that obeys a movement remaining agents. According to the authors of this article, the principle of selfish herd is formalized by the method of steepest descent. Keywords: Natural algorithms, method of steepest descent, algorithm of the Shepherd’s dog, multi-agent systems, cybernetics, programming, biosimilar methods, biological metaphor, bioinspired approach. Contacts: E-mail: evd-ivan@yandex.ru Pp. 10-13.