Network model of spreading of several activity types among complex agents and its applications

The paper describes the main principles of a model that simulates distribution of several types of activity among the agents having internal structure. The potential interaction of agents is given by a weighted graph. The vertices of the graph are heterogeneous automata with infinitely many states, and edges correspond to the influence of vertices on each other. The activity is simulated by the propagation of the integer resource called chips. The vertices exchange chips of m different types along edges. Each edge of the network has m non-negative weights characterizing the capacity of each type. We define the set of set relationships and operations on chip types, so that the vertex can not only change its internal state, but also affect the activity of the network configuration in general. As an example, the model of the threshold interactions in the social network with two types of activity and five types of agents with different activation threshold is described.