Analysis and development of agent architecture for pest control systems

Abstract

While developing the control system process in which robot performing various operations it is essential observance of many limitations. Currently, there are technological capabilities to take into account many factors in the estimating of the large amount of calculations associated with the existing of factors and constraints. It is known that there are many factors occur at random during the execution of operations. To solve the explicit and implicit factors it is necessary to improve the architecture of existing system. In this paper we analysis previously developed MAS, focusing on its architecture and its agents. As the result of analysis there found main indicators of MAS in terms of their structures, actions, parameters, type, number and purpose. Described the agent forming procedure, where it is carries out movements in a predetermined pattern. During the execution of tasks agent is in an environment, where it is meet different obstacles in form, type and level of complexity. Also we considered several obstacles avoidance algorithms. Presented architecture of the agents will be used in the multi-agent based pest control systems, which will be used in agriculture. © 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license.