Action selection in the artificial fish is controlled by its behavior system. The behavior system consists of the habits and mental state of the fish, an intention generator and a set of behavior routines (Fig. ). The behavior system runs continuously within the fish's simulation loop. At each time step the intention generator issues an intention based on the fish's habits, mental state, and incoming sensory information. It then chooses and executes a behavior routine which in turn selects and runs the appropriate motor controllers. It is important to note that the behavior routines are incremental by design. Their job is to get the artificial fish one step closer to fulfilling its intention during the current time step. Moreover, at any given moment in time, there is only one intention or one active behavior in the artificial fish's behavior system. This hypothesis is commonly made by ethologists when analyzing the behavior of fishes, birds and four-legged animals of or below intermediate complexity (e.g. dogs, cats) [Tinbergen1951, Manning1979].
In this chapter, we first list a set of design criteria for effective action selection mechanisms that guides our design of the behavior system. Next we discuss the implementation scheme we choose for the behavior control (or action selection) of the artificial fish, considering especially the overall structure of the different control levels and stages. We then describe in more detail how this is implemented via the intention generator and the behavior routines. In addition, we explain how some of the primitive behaviors, such as collision avoidance and moving target pursuit, as well as some of the more complex, motivational behaviors, such as schooling and mating, are generated. We also introduce the different types of artificial fishes that we have implemented to date. Finally, we present results and discuss possible improvements to our design.
Figure: The behavior system in an artificial fish.
|Xiaoyuan Tu||January 1996|