Modelling and Simulation of Adaptive Multi-Agent Systems with Stochastic Nets-within-Nets

This study centers on self-adapting multi-agent systems modeled utilizing the SONAR framework. Our key focus is on forecasting the costs and benefits of adaptation during execution within the MAPE loop (monitor, analyse, plan, and execute). Analyzing these adaptation processes is intricate due to SONAR enabling second-order activities, such as structural adaptation involving agent interaction protocols or the organizational network itself. We forecast these dynamic processes using a stochastic run-time model (e.g., the environment has a stochastic representation). Since SONAR is conceptualized with HORNETS (a nets-within-nets formalism), we necessitate “probabilistic” HORNETS. To illustrate our approach’s effectiveness, we showcase a small case study of a self-modifying MAS organization and provide an analysis of adaptation dynamics.