PriceCast fuel: Agent based fuel pricing

Abstract

Setting the right price at a gas station is a complex task involving numerous parameters. By using a hybrid agent architecture based on BDI and ANN we can model a gas station agent that can learn to model its consumers. The gas station agent can then use the learning from it’s consumer behavior to detect anomalies in the environment and autonomously set its own price to influence the consumer and thereby optimize e.g. gross margin without sacrificing volume.