Development of a fuzzy logic-embedded system dynamics model to simulate complex socio-ecological systems

To respond to the growing challenges posed by adverse environmental impacts and climate change, there is an increasing need for multidisciplinary and comprehensive research to build sustainable socio-ecological systems (SES). System dynamics (SD) has been widely used as a methodology to meet these needs, but the common practice of oversimplifying or subjectively handling complex relationships among various factors often reduces the reliability of the model. Therefore, the objectives of this study were (1) to develop a methodology for integrating fuzzy logic into the SD model to handle relationships among multiple variables systematically and (2) to validate the effectiveness of the proposed methodology through a case study on a simple SES. The developed methodology encompassed procedures for constructing fuzzy logic, including fuzzification, fuzzy inference, and optimization, on the SD platform. The usefulness of this methodology was tested with a fuzzy-SD model for a rice production system, wherein fuzzy logic was applied to capture variations in rice yield based on temperature conditions. As a result of optimizing the fuzzy-SD model, the rice yield inferred based on two types of temperature factors and eight fuzzy rules closely agreed with historical data (mean absolute percent error = 2.15 %). These results suggest that (1) the methodology proposed in this study can intuitively implement the fuzzy-SD model on the SD platform, and (2) utilizing inference through fuzzy logic can be valuable in minimizing errors within the SD model. Our findings can contribute to enhancing the reliability and utility of SD models for SES research by enabling reasonable inference regarding complex relationships among system components.