The Reliability Study For Power Grids Based On Markov Decision Process

With the development of science and technology, the power grid is developing rapidly and its coverage area is becoming much more widely in our country. Power Grids have shown its importance as the key way of power transmission.However,the maintenance of pwer grid is now getting harder to implement because of the expanding of power grid and the complexity of its structure.The transformer is responsible for the power transmission and power conversion in power grid as the key equipment.During the operation time of power grid, the conventional fixed cycle maintenance policy wastes the cost in the early age of its lifetime, and reduce the reliability in the lately age of the transformers. By capturing the characters of degradation of transformer, an analytical model based on discrete-time Markov control processes is introduced. With the consideration of the differences among the transformers in power grid, a discrete-time Markov control processes is introduced to maximize the reliability of power grid based on the single transformer degradation model. Q-Learning is proposed in this article to obtain the optimal results.Furthermore, the simulation result indicates that the two models are feasible and economical.By capturing the characters of transformers stoppage,a optimizational model based on discrete-time Markov control processes model is proposed with the Consideration of both reliability and economic. Combining the concept of Q-learning and simulated annealing is proposed in this article to get the optimal maintenance policy.