The probability of state transition from st to st+1 when the agent chose action at was
The task of the agents is to find the global maxima of NK fitness landscapes and the agents cooperate by broadcasting messages informing on their fitness to the group.
The primary novelty of the model consists of the dynamic and adaptive planning capability to distribute tasks among the agent members of the organization as effectively as possible.
In addition, we extended the algorithms for the case of multiple repeated tasks, where the agent has a different randomly-chosen task to perform each time.
Our first algorithm, state abstraction from demonstration (AfD), uses a small set of human demonstrations of the task the agent must learn to determine a state-space abstraction.
The final experiment used both types of tasks making the agents consensus on assignments for 10 multi-agent tasks and 10 single-agent tasks.
In general, the crucial step in the BS algorithm is the choice of the task that the agent must execute in the next time step.
However, states e9 and e18 cannot be occupied except possibly on the first step of a task if the agent begins at them, because they have no incoming transitions.
Figure 9 shows a simulation of this task and the agent's performance.
The paper gives a system overview and describes the main tasks of agent based transport planning and operation, including the description of an improved headway control system as an extension to ETCS level 3.
