Fuzzy Agent-based Model: A hybrid tool for exploring spatial perception and behavior

Agent-based Model (ABM) is a recent simulation approach that involves reproducing a real world system into a virtual one, where experiments shall be performed [1][2]. In this virtual world, each agent is represented by an independent entity that is capable of acting locally in response to its perception coded behavior and environmental changes. Fuzzy Logic [3], addition, has been used with relative success to handle real world uncertainty and linguistic terms in computational systems. In this paper we present our first attempt to combine ABM with Fuzzy Logic in order to explore human beings´ spatial perception and consequent behavior expressed in terms of land use/occupation. We understand that this process has a straight relation with environmental attributes and individuals´ preferences. Since individuals might be grouped according to their perception and behavior, we can make use of ABM for simulating land use/occupation according to users groups´ spatial distribution. However, a frequent problem in such approach is how to successfully code spatial perception and behavior. Considering that sociologists and anthropologists normally explore and explain human behavior through natural language, we saw an opportunity to solve the problem employing Fuzzy Systems. The idea consists of using fuzzy rules to code each agent (user group) spatial perception, while the inference system is the decision engine behind each agent behavior. The inference system employs perception maps that derive from spatial analysis as the agent input information, whereas the output is a synthesis of the spatial perception. This synthesis is represented as a mental map and stands for the artificial world were the agent lives. The agent´s agenda (behavior) is to seek in the artificial world the place that best fulfills its preferences/needs. A software prototype of this hybrid model, called Fuzzy Agent-based Model (FAM), was implemented to test the proposed approach through a study case based on coastal use/occupation. The results have indicated that FAM can and has great potential to deal with spatial perception and behavior. There is no doubt that FAM can imitate reality as an approximation of the real world functioning rather than a precise model. So, its basic function is to unders- tand tendencies that result from global system behavior. As a new approach, this kind of application has ahead a challenge to consolidate both concepts and methodology.