Artificial life and online flows optimisation in energy networks

In this paper we propose a methodology to optimally manage and online control energy flows over a power network. Such a methodology is essentially based on a artificial life environment. Exploiting some results achieved in the field of evolutionary computing and artificial life environments, the proposed method is intended to combine the ability to select the current best configuration for the network flows with the capability of building an online model for the performance of the network by means of continuous learning the current situation, adapting its internal actions and updating the suggested optimal solution, which controls the process. With the aim to investigate in the future the possibility of a partially distributed control system, we firstly define the concept of energy district. Hence, we formulate the problem of the online optimal flows management in this type of energy power networks and we finally present some results about the application of the evolutionary control to a real benchmark, the network of the "Casaccia" Research Centre, when critic operating conditions are simulated.