Open AccessMethodology for Selection and Placement of Agricultural Crops using Artificial Intelligence
Author(s) -
В. К. Каличкин,
Roman A. Koryakin,
Tatyana A. Luzhnykh,
Vera Riksen,
Anastasia S. Rudneva
Publication year - 2019
Publication title -
international journal of recent technology and engineering
Language(s) - English
Resource type - Journals
ISSN - 2277-3878
DOI - 10.35940/ijrte.d8638.118419
Subject(s) - selection (genetic algorithm) , computer science , control (management) , set (abstract data type) , domain (mathematical analysis) , artificial intelligence , mathematics , agricultural engineering , engineering , mathematical analysis , programming language
The conceptualization of the domain knowledge “selection and placement of a crop” was carried out. The model contains ontological entities – classes and the relationships between classes, presented in the Unified Modeling Language format. The model describes five classes – “Crop”, “Crop biology”, “Conditions for growth and development”, “Control actions”, and “Placement”, as well as a list of characteristics that affect the performance targets of factors. An algorithm based on the control matrix has been proposed. With the participation of an expert, values from zero to one were assigned for seven factors of “Control actions”, based on the assumption that the control actions affect 22 factors of three other classes: “Crop biology”, “Conditions for growth and development”, and “Placement”. A crop selection can be performed by comparing these matrices with each other. A flow diagram for the preparation of decisions for crop cultivation management, which includes 84 control action chains, has been proposed. Each of these chains was also rated from zero to one for conformity with a certain ideal agronomic strategy. According to these estimates, it is possible to obtain the strongest and weakest strategies for the decision-maker. The structural management of crop selection and placement does not include changing of the models that calculate numbers in the final tables for each crop, but rather is the development of criteria for the presence of the class characteristics in the table for the domain knowledge. Adding or removing characteristics for each crop changes ratings and total values. Artificial intelligence identifies a set of strategies according to the algorithm of their evaluation, which was developed with the participation of an expert.