Agent-based modeling methodology for the development of territorial logging systems

The paper examines methodological and practical aspects of designing agent-based models that support decision-making on the development of territorial logging systems. The aim of the study is to design an agent-based modeling methodology to create models for a territorial logging system. Scientific novelty and significance of the research consist in the creation of specialized approaches to designing logging systems models, in which we elaborate on creating a spatial network, possibilities of its integration with geoinformation systems, ensuring the possibility of adaptation to a service-based approach in the formation of elements, enabling the formation of agents’ behavior in terms of using spatial elements of the model. We consider tasks related to the development of territorial logging systems in Russia, including the creation of an effective transport and logistics network. We analyze the toolkit used to solve the abovementioned tasks. Most studies have formulated the same goal - to reduce the total operating costs of harvesting wood. In this regard, agent-based modeling can claim to be a significant tool for solving this task. The main problem is lack of a methodological basis for building models; therefore, so far it is premature to talk about the possibility of creating a unified methodology, the list of tasks to be addressed is often endless. At the same time, it is possible to narrow the range of issues at hand by focusing on individual subject areas. Thus, we analyze existing approaches to the creation of agent-based systems and formulate our own approach to the creation of agent-based models of territorial logging systems. We put forward an algorithm of specific steps and stages to design and implement agent-based models. It includes creating a contextual diagram of the simulated system, a methodology to form a conceptual and functional structure of the model that is invariant to the tools of agent-based modeling. We consider constructing a spatial environment for models by integrating them with geoinformation systems. At the moment, the concept of an agent-based logging model has been created for the territory of Babushkinsky Municipal District of the Vologda Region, the main aspects of its implementation have been worked out in AnyLogic modeling environment. In order to desing a full-fledged model, there must be interest on the part of those who can provide actual data.