A MODEL FOR PROCESSING APPLICATIONS AND DISTRIBUTING TASKS FOR A ROBOTIC WAREHOUSE

Abstract

The purpose of this work is - development of a model for processing applications and distributing tasks between robots that serve a robotic warehouse. This research is relevant due to increase number of warehouse space, the appearance of stores without buyers (darkstores) and the popularization of purchases through the Internet, which requires the involvement of robots to solve transport problems when arranging orders. To achieve this goal, in this work solves the problem of conceptual representation of a robotic warehouse in the form of a queuing system, which allows using its quality indicators to improve transport processes. Models of the control system of a single robot and processing of orders are presented in the form of finite state machine, which simplifies model experiments and further implementation in onboard robot computers. Propose the criterion for evaluating the duration of the execution of orders by robots, including several types and positions of product in the order, is proposed, which allows processing one order by several robots at the same time. The route of each robot is represented by a set of sections of the path between the collection points of individual products, described in the form of ordered permutations. Such representation made it possible to define a system of inequalities, on the basis of which routes of several robots for processing one order are formed. Algorithms for the distribution of tasks for the robot's onboard computer and the central warehouse server have been developed. The greatest computational load lies on the server because all possible permutations for each order are calculated there. Experimental researches on the simulation model have shown the high efficiency of the developed models and algorithms.