Reinforcement learning solution to economic dispatch using pursuit algorithm

Reinforcement learning (RL) algorithms are powerful tools that can be used to solve multi stage decision making problem. In this paper, we view Economic Dispatch (ED) problem as an n stage decision making problem and propose a novel RL algorithm which uses pursuit algorithm for making decisions at each stage during the learning process. Even though many soft computing techniques like simulated annealing, genetic algorithm and evolutionary programming have been applied to ED, they require searching for the optimal solution corresponding to each demand. In RL approach, once learning phase is over, we can find optimal dispatch for any load from a lookup table. One important issue in RL algorithm is striking a balance between exploration and exploitation during the learning phase. Here we propose to use an efficient algorithm called pursuit algorithm from theory of learning automata for balancing the exploration and exploitation during the learning phase.