Verification Validation Of An Agent-Based Forest Fire Simulation Model

摘要

In this paper, we present the verification and validation of an agent-based model of forest fires. We use a combination of a Virtual Overlay Multi-Agent System (VOMAS) validation scheme with Fire Weather Index (FWI) to validate the forest fire Simulation. FWI is based on decades of real forest fire data and is now regarded as a standard index for fire probability with wide usage across Canada, New Zealand and Australia. VOMAS approach allows for flexible validation of agent-based simulation models. In the current work, it is used in the form of a simulation of a randomly deployed Wireless Sensor Network for forest monitoring. Here, each virtual "sensor" agent uses FWI to calculate fire probability and compares it with the simulation model. VOMAS verification and validation methodology for agent-based models allows for interactive design of Agent-Based Models involving both the Simulation Specialists as well as the Subject Matter Experts. The presented simulation model also uses weather parameters such as wind speed, rain, snow to calculate Indices such as Fire Weather Index (FWI), Build Up Index (BUI) and Initial Spread Index (ISI) in real time. We also study the effects of fires on the life of simulated VOMAS sensors. Using extensive simulations, we demonstrate the effectiveness and ease of use of VOMAS based Validation.